Amino Acid Metabolism in the Inflammatory Niche

AMINO ACID METABOLISM IN THE INFLAMMATORY NICHE

Alice Chun-yin Wang

Department of Haematology Faculty of Medicine Imperial College London

A thesis submitted for the degree of Doctor of Philosophy in Imperial College London

2015

將這份榮耀獻給我的爸爸媽媽

我愛你們

DECLARATION OF ORIGINALITY

The research described in this thesis is the sole and original work of the author. Contributions by other persons have been acknowledged where necessary.

‘The copyright of this thesis rests with the author and is made available under a Creative Commons Attribution Non-Commercial No Derivatives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that they attribute it, that they do not use it for commercial purposes and that they do not alter, transform or build upon it. For any reuse or redistribution, researchers must make clear to others the licence terms of this work’

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my supervisor Prof. Francesco Dazzi for his guidance to expand my scientific knowledge and make my PhD experience stimulating. He has taught me, both consciously and unconsciously, how to grow as a research scientist. The members of the Dazzi group have contributed immensely to my personal and professional time at Imperial. Special thanks to Ilaria Marigo and Cristina Trento, my Italian sisters, who offer endless support and have faith in me even during the tough times in my PhD pursuit. I would also like to thank Antonio Galleu, Yuan-Tsung Li and Luigi Dolcetti, for sharing their enthusiasm for both science and life with me. Other past and present colleagues that I have had pleasure to work with are Ling Weng, Andrea Guerra, Luciene Lopes and Monica Beato Coelho. I am also grateful to George Nteliopoulos for his support and technical advices. I owe a dept of appreciation to Dr. Mark Birrell, Prof. Vincenzo Bronte, Prof. Jorge Caamano, Dr. Neil Rogers, for providing me important research tools.

I am hugely grateful for a few friends who encourage me all the time and their love have constructed a second home for me in London. I thank Kevin, Chien-nien, Wei Wei, Ellen, Carolyn, Isabella, Rebecca, Vincent and Jack for this. Last, but not least, I would like to thank my parents for all their unconditional love, support and encouragement. Thank you.

ABSTRACT

Stroma and parenchyma represent the supportive and functional components in every organ of the body, respectively. Beyond their ability to produce structural support and to differentiate into tissues of mesodermal origin, mesenchymal stromal cells (MSC) exhibit potent immunomodulatory properties. Such a function requires an activation step (‘licensing signal’) provided by the inflammatory microenvironment to which MSC are exposed. My results have attributed immunosuppressive effects of MSC to essential amino acid (EAA) deprivation. Amongst the EAA consuming enzymes examined, blocking nitric oxide synthase 2 (NOS2) and histidine decarboxylase (HDC) both resulted in impaired anti-proliferative activity of MSC, while NOS2 appeared to be a more prominent effector. My results have also demonstrated that TNF-α and IFN- γ differently account for NOS2 and HDC up-regulation, respectively. Furthermore, MyD88 and NF-ĸB were identified as upstream mediators for initiating NOS2 production. The role of TNF-α and NOS2 in MSC-mediated immunosuppression was assessed in vivo using a murine model of peritonitis. MSC treatment remarkably reduced the local inflammatory response during acute peritoneal inflammation. Nevertheless, both Nos2-/- and Tnfr1/r2-/- MSC delivered similar effects compared to WT MSC, indicating the presence of other complementary mechanisms in MSC- mediated immunosuppression in vivo. In addition to their immunomodulatory properties, MSC are fundamental in regulating self-renewal and differentiation of haematopoietic stem cells (HSC). MSC protect HSC from potential damage by maintaining their quiescence. My results have revealed that the ability of MSC to enhance the quiescence of HSC was associated with cell-cycle arrest induced by NOS2. As striking parallels exist between the normal and malignant stem cell niche, I investigated the ability of MSC to protect haematopoietic malignant cells from chemotherapy-induced apoptosis. MSC were observed to confer protection from etoposide- induced necrosis of EL4 cells, possibly due to their ability to suppress EL4 proliferation. Collectively, my results have demonstrated the role of MSC across the fields of immunomodulation, niche-supporting and anti-apoptotic effects.

TABLE OF CONTENTS

AMINO ACID IN THE INFLAMMATORY NICHE…………………………………. 1 DECLARATION OF ORIGINALITY…………………………………………………. 3 ACKNOWLEDGMENTS……………………………………………………………… 4 ABSTRACT……………………………………………………………………………. 5 TABLE OF CONTENTS………………………………………………….……………. 6 LIST OF FIGURES AND TABLES………………………………………………….… 8 ABBREVIATIONS…………………………………………………………………….. 10

1 INTRODUCTION…………………………………………………………………... 14 1.1 Mesenchymal stromal cells (MSC)………………………………………...... 16 1.1.1 Source and ontogeny of MSC…………………………………………...... 17 1.1.2 Phenotype of MSC…………………………………………………………… 18 1.2 Role of MSC in inflammation………………………………………………….… 20 1.2.1 Inflammation determines MSC function…………………………………….. 21 1.2.2 Regulation of immune cells by MSC………………………………………… 22 1.2.2.1 Recruitment of immune cells………………………………………….. 22 1.2.2.2 Re-education of immune cell functions……………………………...... 25 1.2.2.3 Inhibition of immune cell proliferation……………………………….. 31 1.2.3 EAA deprivation in the regulation of immune responses…………………… 34 1.2.3.1 Arginine metabolism by ARG and NOS…………..………………….. 36 1.2.3.2 Tryptophan metabolism by IDO………………………………………. 45 1.2.3.3 Histidine metabolism by histidine decarboxylase (HDC)…………….. 47 1.2.3.4 Phenylalanine metabolism by Interleukin-4 Induced Gene 1 (IL4I1)… 50 1.3 Role of MSC in haematopoiesis………………………………………………….. 51 1.3.1 The endosteal niche………………………………………………………….. 54 1.3.2 The vascular niche…………………………………………………………… 55 1.3.3 MSC in the haematopoietic niche……………………………………………. 57 1.3.4 Role of MSC in HSC transplantation………………………………………... 59 1.3.5 Contribution of MSC to haematopoietic malignancies……………………… 61 1.4 Hypothesis and aims………………………………………………………...….... 63 2 MATERIALS AND METHODS…………………………………………………... 64 2.1 Chemicals and Reagents………………………………………………………….. 64 2.2 Animals…………………………………………………………………………… 64 2.3 Cells and cell culture……………………………………………………………... 65 2.3.1 Generation of MSC…………………………………………………………... 65 2.3.2 Preparation of splenocytes…………………………………………………… 65 2.3.3 Cell culture………………………………………………………..………….. 65 2.3.4. Co-culture………………………………………………………..………….. 66 2.4 PCR………………………………………………………..……………………… 67 2.5 Real-time PCR ………………………………………………………..………….. 68 2.6 Western blot………………………………………………………..……………... 68 2.7 Confocal……………………………………………………..……………………. 69 2.8 Immunosuppressive assay………………………………………………………... 70

2.9 NO quantification………………………………………………………..……….. 70 2.10 Cell viability………………………………………………………..…………… 70 2.11 Quantification of HY-primed T cells stimulated in vivo………………………... 71 2.12 Thioglycollate (TG)-induced peritonitis ………………………………………... 72 2.13 FACS analysis ………………………………………………………………….. 72 2.14 Statistical analysis……………………………………………………………….. 73 3 THE ROLE OF EAA METABOLISM IN MSC-MEDIATED IMMUNOSUPPRESSION………………………………………………………………. 74 3.1 Introduction……………………………………………………………………….. 74 3.2 Results…………………………………………………………………………….. 76 3.2.1 Characterisation of isolated MSC……………………………………………. 76 3.2.2 MSC express a selective pattern of EAA-consuming enzymes in response to activated splenocytes……………………………………………………….... 78 3.2.3 Regulation of MSC immunosuppressive functions by inhibitors of EAA consuming enzymes……………………………………………………...... 81 3.2.4 Cytokine priming selectively skews EAA-consuming enzyme expression in MSC………………………………………………………………………….. 87 3.2.5 NOS2 deficiency leads to the impairment of the immunosuppressive functions of MSC………………………………………...... 93 3.2.6 Eliminating the effect of TNF-α partially impaired the immunosuppressive activities of MSC……………………………………………………...... 99 3.2.7 The role of NF-κB as a key upstream mediator to induce NOS2 expression in MSC...... 105 3.2.8 The effect of MyD88 in regulating NOS2 expression in MSC…………….... 110 3.3 Discussion…………………………………………...... 114 4 THE EFFECT OF MSC ON HAEMATOPOIETIC CELLS: CELL CYCLE REGULATION AND APOPTOSIS……………………………………………….. 122 4.1 Introduction……………………………………………………...………...... 122 4.2 Results……………………………………………………...………...... 123 4.2.1. NOS2 plays a fundamental role in the MSC mediated cell-cycle arrest of activated splenocytes…………...... 123 4.2.2 MSC induced haematopoietic stem/progenitor cells (HSPC) cell-cycle arrest in vitro………………………………………………………………………... 126 4.2.3 Effect of MSC on tumour cell proliferation………………………………….. 132 4.3 Discussion………………………………………………………………………… 138 5 THE FUNCTION OF NOS IN THE MSC-MEDIATED IMMUNE RESPONSE DURING INFLAMMATION ……………………………………………………... 142 5.1 Introduction…………………………………...………...... 142 5.2 Results……………………………...………...... 144 5.2.1 HY responses in mice immunised with male splenocytes …………………... 144 5.2.2 MSC markedly reduced the TG-induced inflammatory infiltration…………. 147 5.2.3. Nos2-/-MSC regulate leukocyte infiltration differently during inflammation 148 5.2.4. TNF-α plays a role in suppressing the percentage of leukocyte infiltrate during inflammation……...………...... 152 5.3 Discussion………………………...………...... 156 6 GENERAL DISCUSSION ……...………...... 160 7 REFERENCES……………………………………………………………………… 166

LIST OF FIGURES AND TABLES

Figure 1.1. Stem cell niche…………………………………………………………… 15 Figure 1.2. MSC control the recruitment of immune cells. ………………………….. 24 Figure 1.3. MSC regulate the differentiation and phenotypic shifts of immune cells... 30 Figure 1.4. MSC inhibit the proliferation of immune cells…………………………... 33 Figure 1.5. Schematic overview of EAA metabolic pathways involved in immune regulation………………………………………………………………… 35 Figure 1.6. NO-mediated signalling pathways in immunomodulation………………. 41 Figure 1.7. Model of HSC niches in the bone marrow……………………………….. 53 Figure 3.1. MSC exhibit dose-dependent anti-proliferative activity…………………. 77 Figure 3.2. EAA consuming enzymes in MSC are differentially regulated by Con A- activated splenocytes…………………………………………………….. 80 Figure 3.3. The dose of NOS2, HDC, and ARG inhibitors applied in the anti- proliferative assay is not cytotoxic………………………………………. 83 Figure 3.4. NOS2 inhibition by 1400W blocks the anti-proliferative activity of MSC………………………………………………………………………. 84 Figure 3.5. HDC inhibition by α-methyl-DL-histidine dihydrochloride counteracts the anti-proliferative activity of MSC…………………………………… 85 Figure 3.6. ARG inhibition by BEC has no effect on the anti-proliferative activity of MSC……………………………………………………………………… 86 Figure 3.7. Dose-dependent effects of TNF-α and IFN-γ on Nos2 and Hdc mRNA expression levels in MSC………………………………………………... 88 Figure 3.8. Profile of EAA consuming enzyme expression in MSC isolated from C57BL/6 mice………...... 91 Figure 3.9. Profile of EAA consuming enzyme expression differs in MSC generated from BALB/c mice……………………………………………………….. 92 Figure 3.10. The absence of NOS2 after cytokine stimulation confirmed the deficiency of NOS2 in Nos2-/- MSC……………………………………... 93 Figure 3.11. Confocal analysis of NOS2 expression in MSC after stimulation by inflammatory cytokines...... 94 Figure 3.12. Deficiency in NOS2 eliminates the anti-proliferative activity of MSC, whereas the presence of HDC induced by IFN-γ partially reverses the effect.…………… 96 Figure 3.13. Abolished NO production is observed in Nos2-/- MSC compared to WT MSC after stimulation…………………………………………………… 98 Figure 3.14. Tnfr1/r2-/- MSC express different patterns of EAA consuming enzymes compared to WT MSC in response to inflammatory cytokines and Con A-activated splenocytes………………………………………………….. 100 Figure 3.15. Inflammation-induced NOS2 upregulation is impaired in Tnfr1/r2-/- MSC………..…………………………………………………………….. 101 Figure 3.16. Inflammation-induced NO production is impaired in Tnfr1/r2-/- MSC…... 102 Figure 3.17. TNF-α plays a crucial role in the anti-proliferative activity of MSC…….. 104 Figure 3.18. Effect of NF-κB inhibitor, BAY11-7085, on TNF-α induced nuclear translocation of p65 in MSC……………………………………………... 107

Figure 3.19. NF-κB signalling is essential for NOS2 expression in response to Con A- 107 activated splenocytes…………………………………...………………... Figure 3.20. NO production is suppressed by NF-κB inhibition………………………. 108 Figure 3.21. NF-κB inhibition impairs the anti-proliferative ability of MSC…………. 109 Figure 3.22. Genotypic analysis of Myd88-/- MSC…………………………………...... 112 Figure 3.23. NOS2 expression is impaired in Myd88-/- MSC in response to inflammatory cytokines and TLR4 ligand…………………………..…… 112 Figure 3.24. Myd88-/- MSC exhibit reduced anti-proliferative activity………………... 113 Figure 3.25. EAA metabolic pathways regulated by MSC……………………………. 121 Figure 4.1. NOS2 deficiency, but not TNFR1/R2 deficiency, impairs the ability of MSC to induce the cell-cycle arrest of activated splenocytes…………… 125 Figure 4.2. MSC regulate the cell-cycle distribution of HSPC differently depending on cell contact……...... 128 Figure 4.3. MSC maintain the cell-cycle distribution of HSPC in response to G-CSF stimulation………...... 131 Figure 4.4. MSC exhibit dose-dependent anti-proliferative activity on tumour cell lines……………………………………………………...……………….. 133 Figure 4.5. MSC is effective at preventing etoposide-induced apoptosis……………. 135 Figure 4.6. The effect of MSC on starvation-induced apoptosis……………………... 137 Figure 5.1. Leukocyte kinetics during TG-induced peritoneal inflammation………... 143 Figure 5.2. Quantification of HY-primed T cells stimulated in vivo…………………. 145 Figure 5.3. Effect of MSC on HY priming in vivo...... 146 Figure 5.4. Infiltrated peritoneal cells in response to TG-induced peritonitis………... 147 Figure 5.5. NOS2 had limited impact on the immunosuppressive effect of MSC in TG-induced peritonitis…………………………………………………… 151 Figure 5.6. Deficiency of TNFR1/R2 modulates the immunosuppressive effect of MSC in TG-induced peritonitis………………………………………….. 155 Table 2.1. List of monoclonal antibodies for FACS analysis……………………….. 73

ABBREVIATIONS

AGM Aorta Gonad Mesonephros AHR Aryl Hydrocarbon Receptor AIF Apoptosis Inducing Factor ALL Acute Lymphoblastic Leukaemia AML Acute Myelogenous Leukaemia Ang-1 Angiopoietin-1 AP-1 Activator Protein-1 APC Antigen Presenting Cell ARG1 Arginase 1 bFGFR basic Fibroblast Growth Factor Receptor bHLH basic Helix-Loop-Helix BM Bone Marrow BMP4 Bone Morphogenetic Protein 4 BMPRIA Bone Morphogenetic Protein Receptor Type IA Bphs Bordetella pertussis-induced histamine sensitisation cAMP cyclic Adenosine MonoPhosphate CAR (C-X-C motif) ligand 12 Abundant Reticular CCL CC Chemokine Ligand CCR C-C Chemokine Receptor CDK Cyclin-Dependent Kinase cGMP Cyclic Guanosine MonoPhosphate CLL Chronic Lymphocytic Leukaemia CML Chronic Myeloid Leukaemia CNS Central Nervous System CO Carbon Monoxide Con A Concavalin A COX-2 Cyclooxygenase-2 CTLA4 Cytotoxic T Lymphocyte-associated Antigen 4 CX3CR (C-X3-C) Receptor CXCL (C-X-C motif) Ligand CXCR (C-X-C motif) Receptor DAMP Damage-Associated Molecular Pattern DC Dendritic Cell EAA Essential Amino Acid EAE Experimental Allergic Encephalomyelitis EAU Experimental Autoimmune Uveitis ECM Extracellular Matrix EGFR Epidermal Growth Factor Receptor ESC Embryonic Stem Cells FADD Fas-Associated Death Domain FasL Fas Ligand FGF Fibroblast Growth Factor FL Fms-like tyrosine kinase 3 Ligand

FLT-3L Fms-Like Tyrosine Kinase 3 Gal-1 Galectin-1 GAPDH Glyceraldehyde-3-Phosphate Dehydrogenase GCN2 General Control Nonderepressible 2 G-CSF Granulocyte-Colony Stimulating Factor GM-CSF Granulocyte-Macrophage Colony-Stimulating Factor GvHD Graft-versus-Host-Disease HDC Histidine Decarboxylase HGFR Hepatocyte Growth Factor Receptor HLA Human Leukocyte Antigen HIF-1 Hypoxia Inducible Factor-1 HSC Haematopoietic Stem Cells HSCT Haematopoietic Stem Cell Transplantations HSPC Haematopoietic Stem/Progenitor Cells ICAM-1 Intercellular Adhesion Molecule 1 IDO Indoleamine 2,3-dioxygenase IFN Interferon IGF-1 Insulin-like Growth Factor-1 IGFBP2 Insulin-like Growth Factor-Binding Protein 2 IGFR Insulin Growth Factor Receptor IKK Inhibitor κB Kinase IL-1R Interleukin-1 Receptor IL-2 Interleukin-2 IL4I1 Interleukin-4 Induced Gene 1 IRAK Interleukin-1 Receptor Activated Kinase IRF Interferon Regulatory Factor ISCT International Society for Cellular Therapy JAK Janus Kinase KGF Keratinocyte Growth Factor KYN Kyrurenine LAAO L-Amino Acid Oxidase Lepr Leptin receptor LFA-3 Lymphocyte Function-associated Antigen 3 LIF Leukemia Inhibitory Factor LTC-IC Long-Term Culture-Initiating Cell LPS Lipopolysaccharide LSC Leukaemia Stem Cell MAPC Multipotent Adult Progenitor Cell MAPK Mitogen-Activated Protein Kinase MCAM Melanoma Cell Adhesion Molecule MCP-1 Monocyte Chemotactic Protein-1 MDS Myelodyplastic Syndrome MDSC Myeloid Derived Suppressor Cell MEL Murine Erythroleukemia MHC Major Histocompatibility Complex MIF Macrophage migration Inhibitory Factor MIP Macrophage Inflammatory Protein

MK Megakaryocyte MLC Mixed Lymphocyte Cultures MLR Mixed Lymphocyte Reaction MMP Matrix Metalloproteinases MS Multiple Sclerosis MSC Mesenchymal Stromal Cell MSPC Mesenchymal Stem and Progenitor Cell Mx1 Myxovirus resistance-1 MyD88 Myeloid Differentiation factor 88 NF-κB Nuclear Factor kappa B NIK Nuclear Factor kappa B Inducing Kinase NK Natural Killer (cells) NOD Nucleotide Oligomerisation Domain NOD Non-Obese Diabetic NO Nitric Oxide NOS Nitric Oxide Synthase OPN Osteopontin PAMP Pathogen-Associated Molecular Patterns PBMC Peripheral Blood Mononucleated Cells PD-1 Programmed cell Death protein 1 PDGFR α Platelet-Derived Growth Factor Receptor α

PGE2 Prostaglandin E2 PMN Polymorphonuclear leukocyte PRR Pattern-Recognition Receptor PSGL-1 P-Selectin Glycoprotein Ligand-1 PTH Parathyroid Hormone QUIN Quinolinate RAGE Receptors for Advanced Glycation End-products Rb Retinoblastoma RCC Renal Cell Carcinoma RORC Retinoid acid related Orphan Receptor RORγt Retinoid acid receptor-related Orphan Receptor gamma t RIP-1 Receptor Interacting Protein 1 Sca-1 Stem cell antigen-1 SCF Stem Cell Factor SDF-1 Stromal-Derived Factor-1 SEC Sinusoidal Endothelial Cell SLE Systemic Lupus Erythematosus SNO Spindle-shaped N-cadherin+CD45-Osteoblastic SNS Sympathetic Nervous System STAT Signal Transducer and Activator of Transcription TCR T Cell Receptor TG Thioglycollate TGFβRI Transforming Growth Factor-beta Receptor I TIRAP Toll-Interleukin 1 Receptor domain-containing Adapter Protein TNF Tumour Necrosis Factor

TPO Thrombopoietin TRADD Tumour necrosis factor Receptor 1 associated Death Domain protein TRAF6 Tumour necrosis factor Receptor Associated Factor 6 TRAIL Tumour necrosis factor-related apoptosis-inducing ligand Treg Regulatory T cell Trif Toll-Interleukin 1 Receptor domain-containing adaptor- inducing Interferon-beta TSG-6 Tumour necrosis factor-alpha Stimulated Gene/protein 6 VCAM-1 Vascular Cell Adhesion Molecule 1 VEGF Vascular Endothelial Growth Factor VLA-4 Very Late Antigen-4 VSEL Very Small Embryonic-Like stem cell 3-HK 3-Hydroxykynurenine 3-HAA 3-Hydroxyanthranilic Acid β-ME beta-Mercaptoethanol

1 INTRODUCTION

Stem cells are characterised by their ability to self-renew and generate differentiated progeny. A specialised microenvironment that influences the development of stem cells, as well as facilitates their maintenance, is referred to as ‘stem cell niche’. The key components of niche include supportive stromal cells together with the extracellular matrix which they produce (Figure 1.1). Several mechanisms have been identified by which the niche controls stem cell properties. Communication within the niche via secreted factors, which can be extensively produced by stroma, is fundamental for the regulation of stem cell proliferation and fate decisions [1]. Apart from these secreted factors, contact-dependent interactions are also crucial for the regulation of stem cell self-renewal [2]. For instance, in bone marrow (BM), the cell adhesion molecule E- selectin expressed by endothelial cells has been shown to promote HSC proliferation [3]. Extracellular matrix (ECM) in contact with stem cells functions not only as a mechanical support but also as a reservoir for cellular signalling molecules. Accordingly, ECM facilitate integrin/transmembrane receptors connection, which directs a variety of cell behaviours including proliferation, survival and differentiation [4, 5]. Additionally, the spatial relationship and the topographical organization between stem cells and support cells, has been shown to promote asymmetric stem cell divisions in drosophila [6]. Thus, the stem cell niche combines trophic support, structural support, topographic information, and appropriate physiological signals for facilitating stem cell regulation [7].

With these interplays in the niche, the stem cell pool is maintained and the activity of stem cells is preserved to meet the requirement of tissue homeostasis and tissue repair. Tissue homeostasis represents an ongoing balance between cell death and cell replacement under normal conditions. Whilst adult stem cells found in the brain and liver where tissues normally undergo very limited regeneration and turnover, the differentiated offspring of adult stem cells located in blood and epidermis normally have a faster turnover and shorter lifespan. When the tissue integrity is impaired by injury and/or infection, it results in the damage of parenchymal cells and the needs to initiate tissue repair. Tissue repair involves three distinct but overlapping phases, namely inflammation, tissue regeneration, and tissue remodelling.

Stromal cells can participate in tissue homeostasis as a critical component of the niche, controlling cell division and fate determination of stem cells. In particular, mesenchymal stromal cells (MSC) have been shown to exhibit potent immunomodulatory abilities during inflammation.

Figure 1.1. Stem cell niche. This schematic summarises different components of the stem cell niche: stromal cells and extracellular matrix (ECM), soluble factors, and cell adhesion components. The interaction between ECM and stem cells can be mediated by binding via integrin, a transmembrane receptor.

1.1 Mesenchymal stromal cells (MSC)

The terminology MSC has been used to identify a yet undefined heterogeneous population that contain both stem/progenitor cells and ‘accessory’ cells like fibroblasts and perivascular stromal cells. A small proportion of these cell preparations are genuine stem/progenitor cells that, under appropriate conditions, have the ability to differentiate into the three lineages of the mesenchymal pathways including adipocytes, chondrocytes and osteoblasts [8, 9]. There has been great interest in MSC due to their multipotent properties. Some research has extended the ability of MSC to transdifferentiate into endothelial cells [10], neural cells [11, 12] and cells of endodermal origin [13, 14]. However, these findings remain controversial, and in particular, the functional properties of MSC transdifferentiated into ectoderm and/or endoderm remain to be verified.

The vast majority of the cells that has been used for experimental investigations so far and for clinical applications, consists of stromal cells. The term ‘stromal’ initially referred to the supportive function of these cells as a skeletal support of tissues. The characteristics of MSC overlap substantially with the concept of tissue fibroblasts. Although initially isolated from the bone marrow (BM) [15], MSC have been identified and isolated from virtually all tissues including, adipose tissue [16], skin [17], placenta [18] and heart [19]. The International Society for Cellular Therapy (ISCT) has recommended the term ‘mesenchymal stromal cells’ instead of ‘mesenchymal stem cells’ [20] despite the poor tools available to unambiguously characterise these cells. There are concerns due to the lack of globally standardised methods for their isolation, expansion and identification, as well as the diversity of methods used to evaluate their differentiation potential.

Therefore, much effort has been invested both in expanding and phenotypically characterising MSC in vitro. The heterogeneity of MSC has been demonstrated by the presence of clones with different morphologies, proliferative and differentiation capacities [21-23]. Also, MSC are not functionally equivalent regarding their differentiation potential when assayed in vivo [24]. Therefore, understanding the developmental origin that MSC arise from and their functional maturation helps to

16 better characterise the hierarchy of MSC.

1.1.1 Source and ontogeny of MSC

MSC can be isolated from various sources, such as BM [25], adipose tissue [16], dermis [17], teeth [26], placenta [27], and umbilical cord blood [28]. There is no consensus with respect to the differences of MSC generated from different tissues. Their expression of conventional MSC markers and their ability to differentiate are observed to be variable depending on their tissue of origin [29, 30]. Apart from their self-renewal and differentiation properties, there is evidence that a similar functional fashion is shared by all stromal cells [31].

Identifying the ontogeny of MSC during development is fundamental to identify specific markers. The traditional concept that MSC originate from the mesoderm, has been challenged by the observation that the earliest MSC-like cells arise from Sox1+ neuroepithelial cells and that primary MSC are positive for the neuronal marker Nestin [32]. CD146 (melanoma cell adhesion molecule, MCAM) identifies a subset of MSC progenitors characterised by the ability to differentiate into osteocytes. Cultured- expanded CD146+ MSC have been demonstrated to obtain selective function of regulating haematopoietic self-renewal and differentiation, and re-establish the haematopoietic microenvironment in a xenotransplantation model [33]. The overlap between CD146+ cells and Nestin+ cells has yet to be understood.

Throughout ontogeny, mesenchymal progenitors have been associated with features of embryonic stem cells (ESC). Several transcription factors are highly expressed in the pluripotent cells of the inner cell mass, which are critical in regulating the expression of other genes during development. In particular, the expression of transcription factor Oct-4, which is responsible for maintaining the self-renewal and pluripotent properties of stem cells has been associated with the functional properties of mesenchymal progenitors [34]. Oct-4-expressing multipotent adult progenitor cells (MAPC) acquire multi-lineage differentiation potential, and are surmise to be the common progenitors of haematopoietic stem cells (HSC) and MSC [35]. MAPC facilitate haematopoiesis and precede HSC in ontogeny by generating long-term repopulating HSC and the full repertoire of haematopoietic progenitors [36]. Oct-4 expressing very small embryonic- 17 like stem cells (VSEL) represent another population of pluripotent stem cells, which are capable of generating tissue-specific progenitors [37].

Another finding suggests that MSC share a common precursor with HSC and endothelium progenitor cells, termed ‘mesenchymoangioblast’[38]. In line with these studies, the connection between MSC and blood vessels has been established in both normal and pathological states [39, 40]. MSC were found to share similar properties, such as those relating to repair and tissue maintenance, with pericytes, which are the cells that envelop the surface of blood vessels and are in close contact with endothelial cells [39]. MCAM/CD146+ subendothelial MSC interact with developing sinusoids after transplantation and contribute to the establishment of the haematopoietic niche as well as to angiogenesis [33].

1.1.2 Phenotype of MSC

As a result of the heterogeneity in MSC and common features shared with endothelial and fibroblasts, there is no single surface marker that can be used to identify MSC. Therefore, the characterisation of MSC is based on an array of surface markers. Expression of stromal markers, including CD44, CD105, CD73, CD90, and CD106, with the lack of haematopoietic and endothelial markers, such as CD45, CD11b, CD34, CD14, and CD31, has been used to define MSC [41]. At first, STRO-1, a trypsin resistant antigen, was identified as a marker for human MSC. However, the expression of STRO-1 is not consistent during culture and is shared by endothelial progenitors [42, 43].

In rodents, Morikawa and colleagues documented a phenotypic profile with the double expression of platelet-derived growth factor receptor α (PDGFR α) and stem cell antigen-1 (Sca-1) and the absence of CD45 and TER119 (PDGFRα+Sca- 1+CD45−TER119−) representing a subset of MSC with higher differentiation potential. The cells can not only generate colonies at a high frequency in vitro, but also differentiate into hematopoietic niche cells, osteoblasts, and adipocytes after in vivo transplantation [44].

MSC express several cell adhesion molecules of potential importance in cell binding and in the homing of haematopoietic cells, including intercellular adhesion molecule 1 (ICAM-1), ICAM-2, vascular cell adhesion molecule 1 (VCAM-1, also known as CD106), and lymphocyte function-associated antigen 3 (LFA-3). They also express various integrin α and β subunits, which serve as binding units for ECM components in BM [45]. A multitude of growth factor receptors have been identified on the surface of MSC. PDGFR, epidermal growth factor receptor (EGFR), basic fibroblast growth factor receptor (bFGFR), insulin growth factor receptor (IGFR), hepatocyte growth factor receptor (HGFR), and transforming growth factor-β receptor I (TGFβRI) and TGFβRII have all been associated with the self-renewal and differentiation properties of MSC [46]. In addition, MSC have been shown to express a broad range of chemokine receptors, which play a pivotal role in their homing, migration and engraftment abilities. To date, a list of chemokine receptors, including chemokine (C-C motif) receptor 1 (CCR1), CCR3, CCR4-7, CCR9, CCR10, chemokine (C-X-C motif) receptor 4 (CXCR4), CXCR5, CXCR6, and chemokine (C-X3-C) receptor 1 (CX3CR1) have been detected on human MSC [45, 47-49]. The chemokine receptors detected in murine MSC includes CCR2, CCR4-7, CCR9, CXCR3, CXCR4, CXCR6, CX3CR1, and CX3CR6 [50, 51].Notably, the expression of receptors varies due to the heterogenic nature of the MSC population and the time for which MSC are maintained in culture.

1.2 Role of MSC in inflammation

As the first line of host defence, the innate immune system is composed of immune cells, including polymorphonuclear leukocytes (PMN), macrophages, natural killer (NK) cells, dendritic cells (DC), and epithelial cells. These cells constitutively express pattern-recognition receptors (PRR), which recognise pathogen-associated molecular patterns (PAMP) on the surface of bacteria, fungi, and viruses [52-54]. The PRR then trigger rapid and non-specific inflammatory responses against infection. Toll-like receptors (TLR) are one of the PRR expressed prominently amongst antigen-presenting cells (APC), such as DC and macrophages [55, 56]. Activation by TLR ligands recruits Toll/IL-1 receptor (TIR) domain-containing adaptors, such as myeloid differentiation factor 88 (MyD88), to the TLR and initiates signalling events downstream [57]. This results in the activation of nuclear factor-κB (NF-κB), activator protein-1 (AP-1), and interferon regulatory factor (IRF) families of transcription factors, which mediate the gene expression contributing to host defence [52, 58].

In contrast, potent antigen-specific responses and the development of immunological memories are typical features of the adaptive immune response [59]. DC function as a centre to interpret the signals from innate immunity and initiate the adaptive immune responses [60]. As DC undergo maturation and migrate to draining lymph node sites, they become APC, capable of activating naïve T and B lymphocytes [61]. Depending on the subtypes of DC and cytokine profiles they create, DC direct the differentiation of T cells into T-helper 1 (Th1) or T-helper 2 (Th2) phenotypes according to their cytokine profile [62, 63]. Whilst Th1 cells produce pro-inflammatory cytokines like interferon-γ (IFN-γ) and interleukin-2 (IL-2), which promote cell-mediated immunity and phagocyte-dependent inflammation, Th2 cells mainly secrete anti-inflammatory cytokines such as IL-4, IL-5, IL-10, and IL-13; these enhance antibody responses, eosinophil activation and phagocyte-independent protection [64]

Although inflammation is a protective response elicited by tissue damage, it has to be regulated. A number of cellular interactions and molecular mechanisms have been identified that limit the duration of inflammation and convert it in order to facilitate tissue repair. Tissue MSC are at the crossway of such a regulatory enterprise.

1.2.1 Inflammation determines MSC function

During inflammation, cellular components released from apoptotic and necrotic cells or damaged microvasculature not only induce the infiltration of macrophages and neutrophils, but also the recruitment of adaptive immune cells including B cells, CD4+ T cells and CD8+ T cells [65, 66]. Historically, studies on inflammation have focused on the role of TLR signalling and antigen-specific lymphocyte responses. Nevertheless, there is increasing evidence indicating that stromal cells, such as MSC, exert an immunoregulatory effect on cells from both the innate and adapted immune systems during the inflammatory lesion [67-70]. Furthermore, the immune system and stromal cells form a bidirectional relationship: Exposure to the inflammatory milieu composed by tumour necrosis factor-α (TNF-α), IL-1β, IFN-γ, and chemokines shapes the functions of stromal cells. Stromal cells in turn limit inflammation and enhance tissue repair by regulating the production of various immunoregulatory cytokines, cell- mobilization factors and growth factors [71-74]. Thus, together with other stromal cells such as fibroblasts and endothelial cells, MSC regulate the skew of the immune response towards the induction of tolerance.

It should be highlighted here that MSC are not constitutively inhibitory, but they require a ‘licensing signal’ to trigger their immunomodulatory capacity [75-81]. In addition, MSC possess an inherent plasticity with distinct phenotypes, dependent on their exposure to the microenvironment composed of inflammatory molecules and immune cells, including T cells, B cells, and NK cells. MSC are known to express a number of PRR, including TLR, nucleotide oligomerisation domain (NOD) receptors, and receptors for advanced glycation end-products (RAGE) [82-86]. These receptors recognise microbial PAMP and endogenous danger signals released after cell damage, and subsequently activate MSC to deliver different immunomodulatory effects and distinct secretomes. Depending on the TLR stimulated or types of cytokines, MSC can be polarised towards pro- or anti-inflammatory phenotypes, finding a parallel with the M1-M2 macrophage polarisation paradigm. Stimulation with lipopolysaccharide (LPS, a TLR4 agonist), TNF-α or IFN-γ (pro-inflammatory cytokines) induces the production of IL-6, IL-8 or TGF-β generating MSC2-type phenotype, which promotes tissue injury repair activity. However, activation with poly(I:C) (TLR3 agonist), IL-4 or IL-13 (anti-

21 inflammatory cytokines) induces the polarisation of the MSC1-type phenotype, which is associated with the incomplete resolution of the inflammatory response [84, 87, 88].

The concept of MSC polarisation into distinct phenotypes offers a paradigm for interrogating the interaction between MSC and their microenvironment. In autoimmune diseases, such as multiple sclerosis (MS), MSC isolated from patients exhibited a distinct profile with a lesser capacity to inhibit T-cell proliferation and to produce IL‐ 10 and TGF‐β when compared with control MSC [89]. In a mouse lymphoma model, tumour resident MSC expressed elevated levels of CCR2 ligands, which enhances the recruitment of monocyte/macrophages and promotes the growth of transplanted tumours [90]. In addition, the MSC resident in the tumour are able to confer such tumour-enhancing features to naïve MSC via co-culture. Collectively, these data indicate the importance of crosstalk between MSC and the microenvironment.

1.2.2 Regulation of immune cells by MSC

During inflammation, recruited immune cells are modified by a series of coordinated decisions to initiate proliferation, differentiation or apoptosis. MSC have been shown to regulate virtually all types of immune cells, including neutrophils, monocytes, macrophages, T cells, B cells, and NK cells.

1.2.2.1 Recruitment of immune cells

In an early phase following inflammation, danger signals such as microbial components or tissue damage-associated molecular patterns (DAMP) activate tissue-resident macrophages, which subsequently produce neutrophil-recruiting chemokines, such as CXCL1, CXCL2 (macrophage inflammatory protein 2-alpha (MIP2-α)), CXCL3 (MIP2-β), CXCL8, CC chemokine ligand 2 (CCL2; monocyte chemotactic protein-1 (MCP-1)), CCL3 (MIP1-α), CCL4 (MIP1-β) and induce the recruitment of neutrophils [91-94]. The influx of neutrophils then promotes inflammatory monocyte recruitment by producing granule proteins [95-97]. Additionally, the neutrophil granule proteins secreted are capable of activating cells in close proximity, such as endothelial cells, to produce chemokines, thereby generating a milieu that favours the recruitment of

22 monocytes. Ultimately, the resolution process involves signals, which terminate neutrophil infiltration, enhancing the clearance of apoptotic cells. MSC have been shown to modify the recruitment of phagocytic cells including neutrophils, macrophages, and monocytes during inflammation. In response to LPS challenge, MSC increase the recruitment of neutrophils in a CXCL8- and macrophage migration inhibitory factor (MIF)-dependent manner in a transwell system, indicating their role during early phase of bacterial challenge [95]. In vivo, MSC have been reported to induce CCR2-dependent migration of monocytes from the BM into the bloodstream by enhancing the expression of CCL2 following LPS stimulation [98]. In response to tissue injury in a mouse model, MSC have been shown to secret a wide range of growth factors and chemokines, such as vascular endothelial growth factor-α (VEGF-α), insulin-like growth factor-1 (IGF-1), keratinocyte growth factor (KGF), epidermal growth factor (EGF), CCL3, CCL12, CXCL2 (Figure 1.2). Through these molecules, MSC induced the recruitment of macrophages, keratinocytes, and endothelial cells, as well as the proliferation of keratinocytes and endothelial cells, thereby accelerating the wound healing [99].

Alternatively, in rats after transplantation into the surface of deep burn wounds, MSC have been shown to decrease neutrophil and macrophage infiltration, as well as to modulate the cytokine profile found in the wound specimens [100]. This study showed that the production of anti-inflammatory cytokines, including IL-10 and TNF-α stimulated gene/protein 6 (TSG-6) was up-regulated following the administration of MSC, whereas the secretion of pro-inflammatory cytokines, including IL-1, IL-6, and TNF-α, was decreased [100]. Concomitantly, in a mouse model of sepsis, it has been demonstrated that IL-10 secretion by monocytes and macrophages in response to MSC inhibits neutrophil migration from the vasculature [101]. Besides IL-10, TSG-6 has also been shown to mediate a protective effect via its suppression of CXCL8-induced transendothelial migration of neutrophils [102]. MSC have been observed to exert their anti-inflammatory activity through the secretion of TSG-6 in corneal injury [103] and peritonitis [104].

The mobilisation of neutrophils has to be finely balanced; although a high number of circulating neutrophils contributes to efficient pathogen clearance, excessive neutrophil migration into tissues can lead to oxidative damage. Taken together these results 23 suggest that, during the early phases of inflammation, MSC promote the recruitment of neutrophils, monocytes and macrophages, whilst in the later stages, MSC display an anti-inflammatory phenotype, suppressing the migration of phagocytes to prevent excessive tissue damage [105].

Figure 1.2. MSC control the recruitment of immune cells. During the early phase of inflammation, MSC are licensed to secret a range of soluble factors, including (C-X-C motif) chemokine ligand 2 (CXCL2), CXCL8, CC chemokine ligand 2 (CCL2), CCL3, CCL12, macrophage migration inhibitory factor (MIF), vascular endothelial growth factor-α (VEGF-α), insulin-like growth factor-1 (IGF-1), keratinocyte growth factor (KGF), and epidermal growth factor (EGF). The combination of these factors creates a pro-inflammatory milieu, which enhances the recruitment of monocytes, macrophages, neutrophils, endothelial cells, and keratinocytes to mount an efficient immune defence. Instead, during the late phase of inflammation, MSC adopt an anti-inflammatory phenotype and suppress the recruitment of immune cells via producing cytokines, such as IL-10 and TNF-α stimulated gene/protein 6 (TSG-6). Consequently, MSC dampen the inflammation and prevent excessive tissue damage.

1.2.2.2 Re-education of immune cell functions

T cells

Both in vitro and in vivo studies have demonstrated the immunomodulatory effects of MSC on regulating the differentiation of T cells. Firstly, MSC are capable of restoring the imbalance between effector T cells of one or more phenotypes in response to environmental cues. In experimental allergic encephalomyelitis (EAE), which is a mouse model of multiple sclerosis (MS), MSC promoted the generation of Th2 cells, which subsequently induced a Th2-dominant cytokine shift, with increased levels of IL-4 and IL-5, but reduced the production of the Th1/Th17 cell-related cytokines IL- 17, IFN-γ, TNF-α, and IL-12 [106]. MSC have also been observed to delay the development of Th1 cell-mediated autoimmune diabetes mellitus in streptozotocin- treated rats and in non-obese diabetic (NOD) mice [107, 108]. The protective effects generated by MSC shift the pattern of cytokine production from Th1 to Th2, as well as reduce the number of IFN-γ-producing Th1 cells.

Th17 cells, which are considered to be pro-inflammatory and facilitate the host’s defence against extracellular pathogens, represent a subset of T helper cells regulated by MSC in the inflammation site. Human MSC have been found to prevent the in vitro differentiation of naive CD4+ T cells into Th17 cells, as well as the production of IL- 17, IL-22, IFN-γ and TNF-α by mature Th17 cells [109]. Furthermore, in an inflammatory environment, MSC can induce the regulatory phenotype of Th17 cells by promoting the expression of Foxp3, whilst suppressing the Th17 cell-specific transcription factor retinoid acid receptor-related orphan receptor gamma t (RORγt). This phenotype modification enables Th17 cells to exert anti-inflammatory effects and to suppress the proliferation of activated CD4+ T cells. In vivo, more mechanisms have been identified for MSC-mediated Th17 inhibition, including via the secretion of IL- 27, CCL-2, or induction of T-cell anergy [110-112]. In addition to EAE, the ability of MSC to suppress Th17 cell has been documented to ameliorate the disease progression in the experimental models of diabetes, myasthenia gravis, and renal injury [113-115].

Despite the prominent activity of MSC on highly inflammatory T cells, this activity was found to depend on the cytokine profile of the microenvironment. In a murine 25 asthma model, which is considered as a Th2-dominant microenvironment, MSC have been observed to suppress Th2 cytokines (IL-4, IL-5, and IL-13) [116].

In addition to perform their effects directly over Th1/Th17/Th2 cells, another modality by which MSC modulate the plasticity of T cells and produce immunomodulation is via indirect mechanisms, including promotion of regulatory T cells (Treg) and modulation of DC phenotypes.

There are two main subsets of Treg: natural Treg, derived from the thymus and capable of recognising diverse self-antigens; adaptive Treg, derived from naïve peripheral CD4+ T cells following T cell receptor (TCR) activation in the presence of TGF-β and IL-2 [117, 118]. Initially, it was found that, following secondary mixed lymphocyte cultures (MLC), MSC preferentially promoted the differentiation of CD4+ T cells co- expressing CD25 and/or CTLA-4, which are the phenotypic markers of Treg [119]. In line with this observation, MSC were found to promote the induction of CD4+CD25+ Tregs from IL-2 stimulated peripheral blood mononucleated cells (PBMC) [120]. A further study demonstrated that MSC increased the proportion of CD4+CD25+Foxp3+CD127- Treg following cultivation with CD3+CD45RA+ or CD3+CD45RO+ fractions [121].

MSC-mediated induction of Foxp3 and CD25 expression in CD4+ Tregs is contact- dependent, although soluble factors, including TGF-β and prostaglandin E2 (PGE2) can also be involved [122, 123]. MSC-T cell interaction via Jagged-1/ Notch-1 signalling represents a novel contact-dependent mechanism mediating Treg induction. The role of Jagged-1/Notch-1 signalling was first described in a murine model, in which the blockage of Notch-1 inhibited TGF-β-induced Foxp3 expression, reduced the number of Treg, and impaired their immunosuppressive activity both in vitro and in vivo [124]. Human MSC were observed to induce Foxp3+ Treg population via Notch-1 signalling pathway [125]. On the other hand, human leukocyte antigen (HLA)-G5 is another mediator generated by MSC to promote the expansion of CD4+CD25+Foxp3+ Treg [126]. During the mixed lymphocyte reaction (MLR), the generation of CD4+CD25+Foxp3+ Treg induced by MSC was impaired in the presence of anti-HLA- G5 antibody, while the immunosuppression mediated by MSC was diminished.

There is increasing evidence to support the functional relevance of MSC-mediated Treg induction in vivo. In a murine colitis model, MSC ameliorated the severity of colitis by initiating CD4+CD25+FOXP3+ Treg and reducing of inflammatory cytokines [127]. Similarly, MSC have been shown to regulate the suppression of allergen-driven airway inflammation. The importance of Treg generation is well demonstrated in this model, as the therapeutic effects of MSC were found to be lost after Treg depletion [128]. Following transplantation, MSC have been observed to promote the survival of kidney, liver, and heart allotransplantation through their ability to induce Tregs and control the immune responses underlying graft rejection [129-131]. In addition, rapamycin appears to work synergistically with MSC to promote allograft-specific tolerance in a heart allotransplantation model. In this case, the induction of Tregs and tolerogenic DC appears to play a key role [132].

Dendritic cells

After encountering antigens, immature dendritic cells (iDC) become mature dendritic cells (mDC). During the process of uptake, transport, and presentation of antigens, iDC up-regulate class II major histocompatibility complex (MHCII) and T-cell co- stimulatory molecule (CD80, CD86) expression, which could be referred to as a checkpoint for DC maturation [133]. This transition is crucial for initiating adaptive immunity as iDC fail to stimulate T cells effectively [134]. However, iDC or semi- mature phenotypes are considered to be involved in the induction of peripheral tolerance [135].

Several pieces of evidence have demonstrated that MSC inhibit both the generation and maturation of co-cultured DC. At the generation stage, MSC inhibit the differentiation of DC from monocytes by inducing cell-cycle arrest in the G0/G1 phase [136]. At the molecular level, MSC have been shown to decrease profoundly the expression of Cyclin D2 in monocytes. It has also been observed that MSC down-regulate the expression of presentation molecules (HLA-DR and CD1a) and co-stimulatory molecules (CD80, CD86), thus impairing the development of DC [137].

Mechanistically, this effect has been attributed to the MSC-induced secretion of PGE2 and IL-6 [138, 139].

During the maturation stage, DC co-cultured with MSC fail to express regular maturation markers, including MHCII, CD40, and CD86, after exposure to inducers such as LPS or TNF-α [139, 140]. Subsequently, by reducing the expression of MHCII, CD40, and CD86 on mature DC, MSC impair the activation and proliferation of T cells [141, 142]. In addition, MSC skew the phenotype of mature DC to the immature status, as illustrated by the reduced CD83 expression and modulation of cytokine secretion with increased IL-10 and decreased TNF-α production [143] [120]. These DC were found to adopt an immunosuppressive phenotype and induce a Th1-Th2 shift. Furthermore, both in vitro and in vivo studies have demonstrated that the alloantigen tolerance effect of MSC is associated with their ability to alter the tolerogenic phenotype of DC, which induces the generation of Foxp3+ Tregs and leads to tolerance induction [144-146]. Alternatively, it has been demonstrated that MSC affect the ability of DC to prime T cells in vivo by blocking their migration to the draining lymph nodes [147].

The functional significance of MSC in modulating the function of DC and has been evaluated in therapeutic approaches. BM derived MSC from myelodyplastic syndrome (MDS, a haematological disease characterised by the disorder of peripheral blood cytopenias) patients can mediate an inhibitory effect on DC differentiation, which subsequently suppressed IL-12 secretion and T-cell proliferation [148]. BM derived MSC from patients with chronic myeloid leukaemia (CML) have been shown to regulate the differentiation of mature DC into a distinct regulatory phenotype with lowered expression of CD40, CD80, CD83, and CD86. These DC displayed a capacity to inhibit T cell proliferation and induce the anergy of T cells, as well as promote the production of Treg [149].

Macrophages

Macrophages have a prominent part not only in the maintenance of tissue homeostasis, but in the regulation of inflammation and host defences [150]. Different stimuli endow macrophages with two distinct subtypes mirroring the Th1/Th2 nomenclature: the classically activated M1 phenotype, and alternatively activated M2 phenotype [151]. In the presence of LPS and inflammatory cytokines, such as IFN-γ and TNF-α, macrophages are skewed into a ‘classically activated’ M1 phenotype. A characteristic 28 of M1 macrophages is their capacity to produce high levels of oxidative metabolites (e.g., nitric oxide (NO), reactive oxygen intermediates, and superoxide), pro- inflammatory cytokines (e.g., TNF-α, IL-1, IL-6, and IL-23), and chemokines, including CCL5, CXCL9, CXCL10, and CXCL16 [150, 152]. M1 macrophages play essential roles in killing pathogens and tumour cells, but also cause damage to healthy tissues. Conversely, anti-inflammatory cytokines, such as IL-4 and IL-13, induce the differentiation of ‘alternatively activated’ M2 macrophages. M2 macrophages express high levels of the M2 markers, Ym1 and arginase 1 (ARG1). They also express relatively low levels of proinflammatory cytokines, but high levels of anti- inflammatory cytokines (e.g., IL-10 and TGF-β), which leads M2 macrophages to primarily promote angiogenesis and tissue remodelling, whilst suppressing destructive immunity.

It has been demonstrated that resting monocytes co-cultured with human MSC become M2 macrophages producing indoleamine 2,3-dioxygenase (IDO) [153]. In addition, MSC have also been found to induce a phenotype shift from activated M1 macrophages into an M2 regulatory profile with the induced production of IL-10 and the reduced secretion of TNF-α, IL-1β, IL-6, and IL-12 [101, 154, 155]. The ability of MSC to polarise M2 macrophages is closely linked to their secretion of PGE2 as the activation of PGE2-prostaglandin EP2 and EP4 signalling in macrophages leads to IL-10 induction [101]. Macrophage induced secretion of IL-10 is considered to be the key action of MSC in the improved survival in a sepsis model [101]. By inducing an anti- inflammatory M2 phenotype, MSC have also been found to switch the cytokine profile from pro-inflammatory to anti-inflammatory and subsequently to enhance wound repair [154]. Likewise, MSC have been reported to elicit the M2 polarisation of macrophages and to modify the relevant cytokine profiles, which in turn reduced scar tissue formation and enhanced recovery after spinal cord injury [156]. Interestingly, these polarizing effects of MSC on M2 macrophages have been associated with CCL18 secretion, which worked in conjunction with TGF-β and promoted the generation of Tregs, thus amplifying the immunosuppressive effects of MSC [157].

Figure 1.3. MSC regulate the differentiation and phenotypic shifts of immune cells. In a T helper 1 (Th1) dominant microenvironment, MSC convert Th1 cells into Th2 cells, inhibit the differentiation of Th17 cells, and promote the differentiation of regulatory T cells (Treg). In addition, MSC are able to modify the differentiation of monocytes into M2 macrophages, as well as the conversion of macrophages from M1 into M2 phenotype. This polarising effect of MSC on M2 macrophages also promotes the generation of Treg. Also, MSC can inhibit the differentiation of monocytes into mature dendritic cells (DC), and skew mature DC to an immature DC state. Conversely, MSC have been shown to suppress Th2 cytokine production in a Th2 dominant microenvironment.

1.2.2.3 Inhibition of immune cell proliferation

T cells

Another important feature of MSC is their ability to regulate the proliferation of innate and adaptive immune cells. The effects of MSC on T cell suppression have been well characterised in murine studies [158, 159], as well as some in primates, such as, baboons [160] and humans [161-163]. MSC are capable of inhibiting the proliferation of both naïve and memory antigen-specific T cells in a dose-dependent fashion. In addition, the suppression of T cell proliferation is not MHC restricted as it can be affected by both autologous and allogenic MSC.

Experiments performed with transwell membrane indicated that MSC can mediate their immunosuppressive ability through the release of soluble factors. Hepatic growth factor (HGF), TGF-β [163], IL-6 [164], IL-10 [165], IDO [166], NO [167], semaphorin-3A, galectin-1(Gal-1)[168], and Gal-9[169] have been identified as participating in MSC- regulated T cell proliferation. Interestingly, those studies also demonstrated species differences in the mechanisms involved, as human MSC mediate IDO to inhibit T-cell proliferation, whereas murine MSC exert this action via nitric oxide synthase 2 (NOS2). The details of the IDO and NOS2 role in essential amino acid (EAA) deprivation will be discussed more extensively in the next section.

On the other hand, evidence has revealed the involvement of cell contact in MSC- mediated immunosuppression although it may not be indispensable. For instance, the neutralization of CXCR3, a receptor for the T-cell chemokines CXCL9, CXCL10, and CXCL11, recovered T-cell proliferation inhibited by MSC [79]. In line with that study, MSC express high levels of ICAM-1 and VCAM-1, suggesting their need to interact in close proximity with T cells in an inflammatory environment [170]. The absence of these adhesion molecules results in a significant impairment of MSC-mediated immunosuppression. In human MSC, programmed cell death protein 1 (PD-1), also called B7-H1, has been shown to be induced by IFN-γ and participate in T-cell suppression via cell contact [171, 172].

The suppression induced by MSC can be attributed to cell-cycle arrest, which halts T cells in the G1 phase by inhibiting cyclin D2 and up-regulating cyclin dependent kinase inhibitor p27kip1 [173]. It has also been demonstrated that MSC maintain T cells in an anergic state, while this effect can be partially rescued by IL-2 stimulation [112]. As T cell proliferation cannot be fully recovered, it has been proposed that MSC may induce a divisional arrest anergy in T cells [174]. This type of anergy is associated with a combination of CTLA-4 signalling and a failure to progress to cell cycle. Alternatively, MSC can induce clonal anergy in T cells, which is a state of unresponsiveness caused by T-cell recognition of an antigenic peptide-MHC complex, as MSC do not express essential co-stimulatory molecules including CD80 and CD86 [159, 162, 175].

B cells

B cells development in the BM is strictly dependent on close interaction with stromal cells [176]. Murine MSC have been reported to suppress the proliferation of B cells after stimulation from pokeweed mitogen or T-cell dependent stimuli, such as CD40 and IL-4 [172, 173]. Consistent with these findings, human MSC reduced the proliferation of B cells activated with anti-Ig antibodies, soluble CD40 ligand and cytokines [177]. In addition, the same study also suggested that paracrine signals from B cells are essential for MSC to exert suppressive actions [177]. IFN-γ has been identified as an essential stimulation for MSC to reduce the proliferation of B cells, possibly due to its ability to induce IDO expression [75]. In the experimental model of human systemic lupus erythematosus (SLE), allogeneic MSC from BXSB mice inhibited the proliferation, activation and IgG production of B cells, as well as the proliferation of T cells [178]. In fact, similar to T cell suppression, B-cell proliferation is inhibited by MSC via induction of arrest of the cell cycle at the G0/G1 phase.

NK cells

Both IL-2 and IL-15 are capable of inducing the proliferation and cytotoxic activity of resting NK cells. It has been demonstrated that MSC abrogate IL-2 or IL-15 driven proliferation of NK cells, and reduce their cytotoxicity against HLA class I–expressing target cells [75, 179]. Interestingly, pre-activation by IL-2 or IL-15 increased the susceptibility of NK cells to MSC-mediated suppression [179]. Experiments with 32 transwell culture systems have shown that MSC regulate the proliferation of NK cells via soluble factors [180]. In the presence of candidate inhibitors, the blocking of both

IDO and PGE2 were found to restore NK-cell proliferation [181]. Other studies have shown that neutralisation of both PGE2 and TGF-β completely restored the capacity of NK cells for proliferation, demonstrating that these factors, when secreted by MSC, suppress NK-cell activity via different mechanisms [180].

(A) (B)

Figure 1.4. MSC inhibit the proliferation of immune cells. (A) MSC suppress the proliferation of T cells, B cells, and NK cells via multiple pathways, involving soluble factors and cell contact. Numerous soluble factors have been identified, including hepatic growth factor (HGF), transforming growth factor-

β (TGF-β), prostaglandin E2 (PGE2), interleukin 6 (IL-6), IL-10, indoleamine 2,3-dioxygenase (IDO), nitric oxide (NO), semaphorin-3A, galectin 1(Gal-1), and Gal-9. On the other hand, MSC mediate a close interaction with immune cells by expressing high levels of intercellular adhesion molecule 1 (ICAM-1), vascular cell adhesion molecule 1 (VCAM-1), and programmed cell death protein 1 (PD-1). (B) The inhibited proliferation is induced by cell-cycle arrest, which halts T cells in the G1 phase. This is associated with up-regulation of cyclin dependent kinase inhibitor p27kip1 that inhibits the activity of cyclin D2, thus impairing the release of E2F transcription factor.

1.2.3 EAA deprivation in the regulation of immune responses

Several mechanisms have been proposed to participate in MSC-mediated immunosuppression. Of relevance, the ability of MSC to up-regulate EAA consuming enzymes such as IDO and NOS2 seems to be a potential target in establishing immunological tolerance. EAA represent a subgroup of amino acids; cells cannot synthesize them or can only partially synthesize them and so they have to be acquired from the environment [182]. Through the competition of essential nutrients, EAA catabolism defines a survival strategy for hosts to control pathogen invasion. In evolutionary terms, EAA consuming enzymes function as potent mediators in the promotion of an immunological tolerance state. The immunosuppressive action of EAA consuming enzymes relates to two mechanisms: the depletion of EAA or the accumulation of their metabolites. The induction of EAA starvation works in conjunction with the subsequent metabolic products to suppress the proliferation and functioning of immune cells (Figure 1.5). Apart from their presence in MSC, EAA consuming enzymes have also been observed in DC, monocytes, macrophages, and myeloid derived suppressor cells (MDSC) [183-186]. In the following chapter, the characteristics of EAA consuming enzymes will be discussed according to the type of EAA used as substrate.

Figure 1.5. Schematic overview of EAA metabolic pathways involved in immune regulation. EAA catabolising enzymes can modify immune responses by controlling the access to essential amino acids, and by producing metabolic products. Here, I focus on five different types of EAA catabolising enzymes, namely, nitric oxide synthase (NOS), arginase (ARG), histamine decarboxylase (HDC), indoleamine 2,3-dioxygenase (IDO), interleukin-4-induced gene 1 (IL4I1). Arginine is a common substrate for ARG and NOS. There are two isoforms of ARG: ARG1 and ARG2; both are capable of catabolising arginine to ornithine and urea. Three isoforms of NOS have been identified: NOS1, NOS2, NOS3. They can all catabolise arginine into NO and citrulline. In particular, when both ARG1 and NOS2 are activated, they - promote the generation of hydrogen peroxide (H2O2) and peroxynitrite (ONOO ), which have profound effects in immune regulation. HDC serves as the only enzyme that catalyses the formation of histamine from histidine. IDO is the key enzyme that catalyses tryptophan along the kyrurenine pathway. As kyrurenine (KYN) is produced as an intermediate metabolite, it can be metabolised further to 3- hydroxykynurenine (3-HK), 3-hydroxyanthranilic acid (3-HAA), and quinolinate (QUIN) by additional enzymes. IL4I1 is a secreted L-amino acid oxidase, which favours the catabolism of aromatic amino acids, phenylalanine, and produce H2O2 and ammonia.

1.2.3.1 Arginine metabolism by ARG and NOS

Arginine is a non-essential amino acid in adults as it can be synthesized from glutamine, glutamate and proline. However, endogenous synthesis of arginine cannot fully meet the metabolic requirements of children or adults when under catabolic stress, which classifies arginine as a semi-essential or conditionally essential amino acid [187]. Arginine serves as a precursor to a wide range of compounds including urea, polyamines, proteins, creatine, agmatine and NO [188, 189]. It has been shown that several immune cells, such as macrophages, neutrophils, and lymphocytes require arginine to maintain their functionality [190]. Therefore, enzymes controlling most aspects of arginine metabolism, namely, ARG and NOS, tightly regulate the immune system in physiological and pathophysiological conditions.

ARG

There are two isotypes of ARG expressed in mammals: ARG1 and ARG2. Both are capable of catabolising arginine to ornithine and urea. Apart from similarities in catabolising the same substrate, the two isotypes are distinct in their transcriptional regulation, tissue distribution and subcellular localization [191]. ARG1 (also known as liver-type) is a cytosolic enzyme, expressed predominantly in liver where it regulates the detoxification of ammonia in the urea cycle. ARG2 (also known as kidney-type) is localised in mitochondria and is expressed at low levels in several tissues (e.g. prostate, kidney, brain, small intestine and lactating mammary glands) [192]. ARG1 has been shown to be up-regulated in macrophages by cyclic adenosine monophosphate (cAMP),

LPS, PGE2 and anti-inflammatory cytokines, including IL-4, IL-10, and IL-13 [193- 197]. In dendritic cells, the expression of ARG1 is induced by IL-4 and IL-10 [197]. ARG1 is expressed constitutively in granulocytes, but is unresponsive to either pro- inflammatory or anti-inflammatory stimuli [198]. In contrast to ARG1, the expression and role of ARG2 in the immune system are less understood. In macrophages, ARG2 can be induced following the stimulation of IRF-3, cAMP, LPS, and PGE2 [199-201]. Interestingly, although ARG2 seems to be less sensitive to pro-inflammatory and anti- inflammatory cytokines, IL-10 was found to synergise with isoproterenol to increase ARG2 expression in macrophages [202].

Two arginine metabolic pathways, which contribute to immunomodulatory effects, have been identified after the local starvation of arginine: (1) down-regulation of the CD3ζ chain and the TCR complex expression; (2) reduced expression of the cyclin D3/ cyclin-dependent kinase 4 (CDK4) complex, which is known to regulate cell cycle progression [203].

Following stimulation by IL-4 and IL-13, murine macrophages induce the expression of ARG1, which then triggers a rapid reduction in arginine. In the absence of arginine, the synthesis of the TCR-associated CD3ζ chain is reduced, causing impaired antigen- specific T cell responses and the diminished proliferation of T cells [204]. In a murine model of lung carcinoma, MDSC isolated from tumours have been demonstrated to produce ARG1, which subsequently inhibited CD3ζ expression and blocked T cell proliferation [205]. In addition, high levels of ARG1 and decreased CD3ζ levels with impaired T cell responses have also been described in patients with metastatic non- small-cell lung cancer and renal cell carcinoma (RCC) [205, 206]. Interestingly, instead of ARG1, only ARG2 has been detected in murine RCC cell lines, in which high levels of the enzyme induce arginine deprivation and suppress CD3ζ expression in T cells [207].

In the absence of arginine, activated T cells are arrested in the G0/G1 phase of the cell cycle. The cell cycle transition is tightly regulated by CDK families, among which,

CDK4 interacts with cyclin D3 to mediate progression through the early G1 phase and into the S phase of cell cycle [208]. This regulation involves the phosphorylation of the retinoblastoma (Rb) protein family [209]. By phosphorylating Rb proteins, the cyclin D3/CDK4 complex induces the release and the nuclear translocation of E2F, thereby permitting cells to proceed into the S phase and initiate DNA replication. Arginine depletion has been shown to block T-cell proliferation via impairment of cyclin D3 and CDK4 [210].

Another predominant effect of ARG activity is associated with the production of ROS and/or RNS through its interaction with NOS2 [211, 212]. The widely known ROS ●- include the superoxide anion (O2 ), hydrogen peroxide (H2O2), and the peroxyl radical (OH●), while common RNS induce NO and peroxynitrite (ONOO-). In the absence of arginine, ARG1 promotes the production of the superoxide anion, which subsequently 37 reacts with NO and results in the generation of peroxynitrite. Peroxynitrite has been demonstrated to induce protein modifications, such as the nitration of tyrosin residues. This can lead to multiple effects, such as dysfunction in IL-2 receptor signalling in T cells, unresponsiveness of CCL2-mediated T-cell infiltration, or induction of apoptotic cell death in T cells [212-214].

Therapeutically, it has been demonstrated that pegylated ARG1 can block the proliferation of T-cell acute lymphoblastic leukaemia in vivo; which correlates with cell-cycle arrest, low expression of cyclin D3 and the induction of tumour cell apoptosis [215]. Administration of pegylated ARG1 or IL-13 treated MDSC, which potently secrete ARG1, has been observed to reduce graft-versus-host-disease (GvHD) lethality in recipient mice. Such effect is associated with diminished CD3ζ expression, impaired T cell activation, as well as reduced pro-inflammatory cytokine production [216].

NOS

Apart from ARG, arginine can also be metabolised by a family of enzymes known as NOS. Three isoforms of NOS have been identified: NOS1 (nNOS), which is mainly expressed in the nervous system; NOS2 (iNOS), which is inducible and expressed primarily in the immune and cardiovascular systems and NOS3 (eNOS), which is expressed in endothelial cells. These isoforms display about 50% sequence identity, and while they are all capable of catabolising arginine into NO and citrulline, they differ in intracellular localisation, regulation, and catalytic properties [188, 217].

NOS1 and NOS3 are constitutively expressed at low levels and their transcription is mediated in a calcium-dependent manner. NOS1 plays a pivotal role not only in modulating physiological functions such as neurotransmittion, neural development and regeneration, but also in a variety of neurological disorders in which excessive production NO results in neural damage [218]. In addition, NO generated by NOS1 in the central nervous system (CNS) has been found to participate in the central control of blood pressure, as inhibition of NOS1 activity in the brain leads to systemic hypertension [219]. Interestingly, the production of NO by NOS1 has a prominent autocrine effect on supporting cytokine-induced maturation of human monocyte- derived DC [220]. NOS3 is mostly expressed by endothelial cells. NO produced by 38

NOS3 has been implicated in the maintenance of systemic blood pressure, vascular remodelling, and angiogenesis [221-223]. NOS3 deficient mice revealed the requirement of NOS3 in facilitating wound healing, possibly due to their effect on enhancing angiogenesis [224]. Furthermore, the enhanced interaction between NOS3 and heat shock protein 90 has been shown to prevent endothelial cell apoptosis induced by angiotensin II [225]. Therapeutically, implantation of MSC overexpressing NOS3 was found to further ameliorate the pulmonary hypertension-related right ventricular impairment and survival [226]. The effects can be attribute to NOS3-induced vasodilation and vascular regeneration. The selective inhibition of NOS3 by the caveolin-1 scaffolding domain has been demonstrated to suppress acute inflammation in vivo, presumably via attenuating NOS3-dependent blood flow and/or permeability changes induced by inflammatory stimuli [227].

In contrast to the constitutive isoforms of NOS, NOS2 functions in a calcium- independent manner and its expression is induced by a variety of extracellular stimuli from bacterial components such as LPS, or pro-inflammatory cytokines, including TNF-α, IFN-γ and IL-1β. In addition, a strong synergism in promoting NOS2 expression was found between the actions of LPS, TNF-α, IFN-γ, and IL-1β [228].

The binding of LPS and IL-1 to the membrane proteins, TLR and IL-1 receptor (IL- 1R), respectively, follows on to a homologous cytoplasmic signalling domain, which is referred as to TIR. Following their activation pathway, both stimuli activate the intracellular signalling cascade via adaptors including IL-1R activated kinase (IRAK) and MyD88, which in turn trigger downstream molecules including TNF receptor– associated factor 6 (TRAF6). Subsequently, the activation initiates signalling pathways including the mitogen-activated protein kinase (MAPK) pathway and the NF-κB pathway. Both of these pathways converge to activate NOS2 transcription [229-231]. TNF-α is recognised by two specific receptors: TNF receptor 1 (TNFR1) and TNF receptor 2 (TNFR2). Signalling via TNFR1 starts with the recruitment of the TNFR1 associated death domain protein (TRADD), which recruits two additional molecules: receptor interacting protein 1 (RIP-1) and TRAF2. Lacking TRADD, TNFR2 interacts directly with TRAF2. Ultimately, via TRAF2, both signalling pathways result in the activation of NF-κB [232].

IFN-γ mediates NOS2 expression via the activation of the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway, which increases the synthesis of the transcription factor, IRF-1 [233]. IRF-1 is a potent mediator for NOS2 transcription, in addition, the interaction between IRF-1 and NF-κB synergistically boosts the NOS2 transcription induced by IFN-γ and LPS [234]. Transcription factors, including STAT1-α and hypoxia inducible factor-1 (HIF-1), were found to regulate NOS2 expression by binding to the promoter sequence of NOS2 [230]. Once activated, NOS2 generates a wide range of NO concentrations, from 10nM to 10μM amounts, for prolonged periods of time [235]. As a potent regulator of NO production, NOS2 exhibits a unique ability to regulate the immune response.

Mechanisms underlying NO-mediated pro- and anti- inflammatory pathways

It has been shown that NO delivers its immunomodulatory functions with both pro- inflammatory and anti-inflammatory actions, either through S-nitrosylation of proteins or by activation of guanylylcyclase, which results in elevated cyclic guanosine monophosphate (cGMP) (Figure 1.6).

Kinases and transcription factors, such as MAPK, JAK/STAT, NF-κB, and AP-1, contain critical cysteine residues that can be modulated by NO and undergo S- nitrosylation. S-nitrosylation of cysteines in these proteins modulates their distinctive functions, demonstrating the biphasic regulatory ability of NO. NO-dependent S- nitrosylation of the p21 subunit has been suggested to induce Ras-mediated NF-κB activation, which contributes to the induction of pro-inflammatory signalling and effector molecules such as cyclooxygenase-2 (COX-2), TNF-α, and IL-6. During this stage, NO-mediated signalling events are associated with the cellular recruitment to the site of inflammation and facilitating host defenses against pathogenic insults [236]. Conversely, NO-mediated S-nitrosylation of inhibitor κB kinase (IKK) suppresses its enzymatic activity, which subsequently prevents NF-κB activation [237]. In addition, NO has also been shown to induce the S-nitrosylation on the p50 subunit of NF-κB reducing its DNA-binding capacity [238]. As a consequence of inhibited NF-κB activity, pro-inflammatory cytokine production is suppressed, thereby contributing to the halting of the inflammatory response.

Whilst mediation through S-nitrosylation is cGMP independent, NO can also mediate anti-inflammatory effects in a cGMP-dependent way [239, 240]. For instance, HO-1 is a microsomal enzyme that cleaves heme-containing proteins and generates biliverdin, free iron and carbon monoxide (CO). By interacting with the heme group of soluble guanylyl cyclase and secreting cGMP, NO has been demonstrated to enhance the production of HO-1 to deliver a cytoprotective effect [240]. In addition, via the activation of guanylyl cyclase, NO functions as an anti-inflammatory mediator suppressing IL-2 induced T-cell proliferation [241].

Figure 1.6. NO-mediated signalling pathways in immunomodulation. Via S-nitrosylation, NO initiates mitogen-activated protein kinase (MAPK) and the Janus kinase/signal transducers and activators of transcription (JAK/STAT) pathways. These pathways further activate transcriptional factors as downstream targets, such as activator protein-1 (AP-1) and nuclear factor κB (NF-κB). Conversely, NO- mediated S-nitrosylation of inhibitor κB kinase (IKK) suppresses its enzymatic activity, which subsequently prevents NF-κB activation. On the other hand, NO can activate the soluble guanylyl cyclase (sCG)-cyclic guanosine monophosphate (cGMP) pathway, which promotes the production of heme oxygenase 1 (HO-1) and carbon monoxide (CO).

NO modulation during innate immunity

As a major component of innate immunity, NK cells are capable of defending against viral infections, parasites, and bacteria without prior sensitization and restriction by MHC antigens. NO has been found to be involved in various aspects of NK cell function [242]. Initially, the genetic deletion of NOS2 in mice was found to abolish NK cell responses including cytotoxic activity and cytokine production in response to Leishmania major infection, suggesting the involvement of NOS2 and NOS2-derived NO in regulating NK cell functions [243]. The lytic activity of NK cells, as well as their production of cytokines, is principally regulated by IL-2 and IL-12. It has been observed that IL-2 activated NK cells exhibit increased NOS2 activity with higher NO production, accompanied by enhanced cytotoxic activity for the lysis of target cells [244, 245]. Application of the NOS inhibitor, L-NAME, significantly suppressed the production of NO and subsequently abrogated the cytotoxic effect of NK cells, thus confirming the essential role of NOS and NO in NK-cell activities [246, 247]. In addition, reduced IFN-γ expression was observed in NK cells after neutralisation of IL- 2 with a concomitant decrease in NO production. This suggests that IL-2 induced IFN- γ production is, at least in part, dependent on NOS2 induction. In the murine model of Leishmania major, the inhibition of NOS2 activity reversed the protective effect of IL- 12, while the cytotoxicity and IFN-γ secretion of NK cells were both abolished [248].

Mast cells are recognised as key players in the initiation of the immune responses during innate immunity, IgE-mediated allergy, and inflammation, due particularly to their proximity to blood vessels and their ability to secrete cytokines [249]. Upon stimulation, mast cells undergo degranulation, rapidly releasing the mediators present within their cytoplasmic granules [250]. Administration of the NO donor, spermine NO, has been shown to inhibit mast cell degranulation in a murine model [251]. In parallel, the NOS inhibitor, L-NAME, is capable of promoting mast cell degranulation [252]. Among the mediators released by mast cells, histamine is associated with increasing vascular permeability and increasing the infiltration of leukocytes. Several studies have demonstrated that NO decreases IgE-induced histamine release from mast cells, which potentially down-regulates vascular permeability [253-255]. In addition, mast cell- dependent leukocyte adhesion to the vascular endothelium is prevented by NO donors [251]. This recruitment effect is reversible as NOS inhibition was found to augment 42 histamine production from activated mast cells. In addition to these observations, NO has been identified as a regulator of other mast cell functions, including protease release [256], cytotoxicity [257], and the production of chemokines and cytokines [258], demonstrating its critical role in immune responses driven by mast cells.

In the innate immune system, the neutrophil is one of the major cell types rapidly recruited to the site of inflammation [259]. During their recruitment, the mobilisation of neutrophils is correlated with their adhesive interactions with the vasculature. In a mouse model of sepsis, a NOS inhibitor was shown to reverse impaired neutrophil migration and improve the survival of mice [260]. Mechanistically, it has been demonstrated that neutrophil migration in response to IL-8 is enhanced by NOS inhibitors [261]. This effect is associated with the generation of microparticles, which express the adhesion molecules L-seletin and P-selectin glycoprotein ligand-1 (PSGL- 1), and are, hence, capable of supporting the migration of neutrophils. In addition, NO has been observed to induce cGMP, which subsequently increases TNF-α production by neutrophils and contributes to inflammatory responses [262]. However, the regulation of neutrophils by NO is obscured by conflicting results showing that NO exerts a biphasic regulatory effect on chemokine production [263-266]. Further investigation revealed that the concentration of NO represents a critical parameter in regulating the recruitment of neutrophils. At low concentrations, NO exhibits pro- inflammatory activity by enhancing the production of chemokines, including CXCL1 and CXCL2, whereas large concentrations of NO become anti-inflammatory by suppressing the production of chemokines and reducing neutrophil infiltration.

NO modulation during adaptive immunity

The immunomodulatory activities regulated by NO on T cells are exerted through the inhibition of T cell proliferation, the induction of T cell subset differentiation, and the regulation of T cell apoptosis [267]. The proliferation of T cells is suppressed by NO through the inhibition of tyrosine phosphorylation, which subsequently down-regulates the signalling intermediates in the IL-2 pathway (e.g. JAK1, JAK3, STAT5, Erk, and Akt) [241, 268].

In addition to regulating genes involved in cellular proliferation and growth, NO exhibits multiple regulatory effects on the differentiation of T cell subsets. Evidence from in vitro studies suggests that NO selectively enhances the stimulation of Th2 cells, but not Th1 cells, and that the expression of NOS2 can be induced by Th1 cytokine IFN-γ, suggesting a role of NO in promoting Th2 differentiation [269]. Experiments carried out in a model of Leishmania major infection have shown that Nos2-/- mice develop enhanced Th1 responses when compared to WT mice [270, 271]. In addition, high concentrations of NO were demonstrated to enhance Th2 differentiation by suppressing IL-12 synthesis [270]. The induction of the Th2 response is absent in Nos2- /- mice, leading to increased severity of experimental autoimmune uveitis (EAU) and EAE [272]. Therefore, this suggests that high levels of NO produced by NOS2 prevent overexpansion of Th1, which is associated with a variety of autoimmune diseases. Conversely, low concentrations of NO selectively promote the differentiation of Th1, but not of Th2 cells. This effect of NO is mediated via the activation of soluble guanylyl cyclase with elevated cGMP [273]. This induced cGMP production subsequently promotes IL‐12Rβ2 expression, which facilitates Th1 differentiation by IL-12- regulated signalling and TCR engagement. The ability of NO to modulate the Th1/Th2 balance is also correlated with its regulatory role in apoptosis. As Th1 cells are more susceptible to apoptosis than Th2 cells, NO is likely to promote Th1 apoptosis at high doses, but to suppress it at low concentrations [274].

NO has been shown to regulate several components of the apoptotic machinery; for example, upon the engagement of the cell death receptor Fas by the Fas ligand (FasL), NO is required to sensitise T cells towards Fas-triggered apoptosis [275]. Apart from this, the generation of mutation patterns in the p53 of mice or human T cells supports the role of p53 in transmitting the pro-apoptotic effects of NO [276]. There is also evidence to suggest that NO triggers T cell apoptosis via regulating expression of Bcl- 2 and apoptosis inducing factor (AIF), or inducing the S-nitrosylation-dependent nuclear translocation and degradation of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [277-279]. However, NO has been demonstrated to protect T cells from apoptosis at low-to-moderate concentrations. Exogenous NO suppresses the apoptosis of T cells potently by interrupting the over-expression of the Fas-associated death domain (FADD) and activating the downstream target caspase-8 [280]. Interestingly, the ability of NO to inhibit caspase activity via S-nitrosylation has been correlated with 44 the suppressive role of NO in T cell proliferation [281]. Apart from mediating a general inhibition of T cell proliferation, NO selectively promotes the functioning and generation of Treg. In a mouse model of schistosomiasis, it has been shown that CD4+CD25+Foxp3+ adaptive Treg interact with macrophages to suppress antigen- driven T cell proliferation via NO production [282]. Interestingly, rat DC transfected with dominant negative form of IKK2 (dnIKK2) have been demonstrated to enhance the generation of Treg [283]. These Treg express high levels of NOS2 and potently suppress T-cell response in vitro and promote kidney allograft survival in vivo.

A unique population of Tregs has been identified whose generation depends on NO action, modulation from p53 and secretion of IL-2 [284]. NO-induced Treg have several similarities to natural Treg as both of their suppressive activities are IL-10 dependent. However, in contrast to Tregs, they fail to suppress Th1 differentiation and functioning, suggesting they constitute a specific subset.

1.2.3.2 Tryptophan metabolism by IDO

IDO is an enzyme that catalyses 90% of tryptophan along the kyrurenine pathway. While kyrurenine (KYN) is produced as an intermediate metabolite, it can be metabolised further to 3-hydroxykynurenine (3-HK), 3-hydroxyanthranilic acid (3- HAA), and quinolinate (QUIN) by additional enzymes.

Within the immune system, IDO is prominently expressed by APC, including macrophages and DC. The activities of IDO is determined by microenvironmental stimuli. Type I (IFN-α/β) and type II (IFN-γ) IFNs can mediate the induction of potent and sustained IDO activity [285, 286]. Additionally, autocrine or paracrine signalling via TGF-β represents another pathway which sustain the activation of IDO and prolong their effects [287]. The functional differences in IDO induced by IFN-γ or TGF-β in plasmacytoid DC (pDC) has been highlighted by an elegant study [288]. IFN-γ- conditioned pDC up-regulated IDO to suppress T cell proliferation and increase T cell apoptosis. On the other hand, TGF-β-conditioned pDC enhanced the generation and maintenance of Treg via IDO. In addition to soluble factors, it has been shown that CD40-CD40L and cytotoxic T lymphocyte-associated antigen 4 (CTLA4)-B7-1

45 interaction on human DC and macrophages significantly induced IDO expression. [289, 290].

Several mechanisms by which IDO modulates immune responses have been proposed, which include tryptophan depletion, generation of downstream tryptophan catabolites, and/or promotion of regulatory phenotypes of APC [288, 291-293]. Both in vitro and in vivo evidence has demonstrated that IDO-mediated depletion of tryptophan from the microenvironment results in halted proliferation of T cells at the G1 phase, while cells become more susceptible to apoptotic stimuli [294]. The mechanism underlying the inhibition of T cell function has been associated with a stress-responsive kinase, general control nonderepressible 2 (GCN2). During tryptophan deprivation, there is an imbalance of tryptophanyl tRNA synthases towards the uncharged forms, which subsequently triggers the activation of GCN2 [295]. It has been demonstrated that IDO- mediated T cell anergy and cell death are GCN2-dependent as GCN2 deficient T cells in mice are resistant to IDO-induced anergy by DC [296]. It is of note that in addition to tryptophan depletion, the activation of GCN2 has also been observed after arginine depletion, suggesting GCN2 to be a common pathway following amino acid deprivation [210].

Tryptophan catabolites, including 3-HAA and QUIN, were found to induce the apoptosis of murine thymocytes and Th1 cells. In particular, Th1 but not Th2 cells rapidly respond to metabolite-induced caspase-8 activation and undergo apoptosis, indicating the potential role of IDO in regulating the Th1/Th2 balance [297]. 3-HK, 3- HAA, and KYN are all effective inhibitors of the proliferation of T cells, while KYN has also been shown to prevent NK cell proliferation [298]. With the aim of targeting potential receptors for individual tryptophan catabolites, the aryl hydrocarbon receptor (AHR) has been identified as a direct target of KYN. AHR belongs to the basic helix- loop-helix (bHLH)/Per-Arnt-Sim transcription factor family and mediates the biological effects of dioxin-like compounds [299, 300]. In a cell-type-specific manner, AHR regulates the transcription of a broad range of genes involved in cell cycle progression, proliferation, cell death and differentiation [301-303]. A prominent role of AHR in immune regulation is its modulatory effect on the differentiation of Th17 cells and Treg [304, 305]. Among the mouse CD4+ T-cell lineage, the expression of AHR is enriched in Th17 cells, while its ligation increases the production of IL-22. However, 46 the activation of AHR has been shown to promote the generation of Treg by controlling Foxp3 expression. It has been observed that tryptophan depletion interacts with downstream catabolites, such as kynurenine, to down-regulate the TCR ζ chain in CD8+ T cells and induce the generation of Treg, so these effects are considered to be linked beyond GCN2 and AHR [306, 307].

IDO was firstly identified as a key molecule in the host defence mechanism against pathogens [285], while further studies revealed its critical role in regulating immunosuppression, such as the prevention of allograft rejection, the induction of protective activities against autoimmunity, and the escape from antitumor immunity [289, 308, 309]. In rodents, up-regulation of IDO in both APC and epithelial cells reduces T cell proliferation and survival, thereby decreasing the severity of GvHD [310]. In contradiction, the frequency of IDO in patients who received allogeneic haematopoietic stem cell transplantations (HSCT) represents a predictive factor for the severity of acute GvHD [311]. In mouse models of autoimmunity, such as EAE, IDO and tryptophan metabolites exert inhibitory effects on Th1 cells and induce a skewing effect via Th2 immune responses, thus reducing inflammation and tissue damage. It has been demonstrated that IDO-deficient mice develop exacerbated EAE with reduced levels of Treg and elevated Th1 and Th17 cell responses [312]. The role of IDO in arthritis has also been assessed and a deficiency of IDO activity increases the severity of collagen-induced arthritis with a higher frequency of Th1 and Th17 cells [313]. In patients with RA, although the synovial DC express higher levels of IDO compared to controls, T cells derived from patients are resistant to IDO-mediated inhibition [314]. However, IDO can regulate multiple mechanisms which contribute to tumour-induced tolerance. These effects can be operated directly by tumour cells via IDO expression in order to inhibit the activity of effector T cells. Alternatively, tumour cells can enhance the expression of IDO in host APC, which consequently promote the activation of Treg and create systemic tolerance [309, 315, 316].

1.2.3.3 Histidine metabolism by histidine decarboxylase (HDC)

Histamine has been identified as a pivotal molecule in a variety of biological and immunological activities. The histamine level is exclusively regulated by histidine

47 decarboxylase (HDC), which catalyses the formation of histamine from histidine [317]. Due to its unique ability, HDC is a specific marker for the biosynthesis of histamine.

In mast cells and basophils, HDC acts as one of the major effector molecules in hypersensitivity reactions, in which the enzyme is constitutively expressed for histamine synthesis and storage [318]. In other cell types including monocytes, macrophages [319], DC [320], neutrophils [321] and T cells [322], the expression of HDC is triggered by inflammatory stimuli, and the histamine is secreted immediately after its production. In response to histamine four differential expressions of receptors,

H1R, H2R, H3R, H4R, have been so far identified. The expression of H1R and H2R is ubiquitous, whereas H3R and H4R expression is restricted to specialised cell populations in the brain and hematopoietic system, respectively.

Via specific histamine receptors, the immune system is differentially regulated according to the location of the inflammation and the type of stimulus for the inflammation. For instance, histamine was observed to exert distinct effects on Th1 and

Th2 subsets, which can be correlated with receptor expression as the H1R is predominant in Th1 cells, whereas the H2R is preferentially expressed by Th2 cells

[323]. In support of this observation, H1R-deficient mice have impaired IFN-γ production, but increased secretion of Th2 cytokines (IL-4 and IL-13) in comparison with WT controls. In H2R-deficient mice, the production of both Th1 and Th2 cytokines is up-regulated. Therefore, histamine can act as an immunostimulant enhancing the Th1 response by triggering H1R, whereas it leads to immunosuppression via H2R, negatively regulating Th1 and Th2 responses. The H4R also modulates cytokine production during the integration of Th1/Th2 responses. A defect in the H4R in T cells dampens Th2 responses, including the production of Th2 cytokines IL-4, IL-5, and IL-13 following antigen stimulation ex vivo [324].

Interestingly, histamine has been shown to modulate cytokine secretion and T cell activation through regulating NO production [325]. In HDC-/- mice, the absence of histamine led to increased IFN-γ production of splenocytes, which was associated with increased NO secretion. Also, the concentration of cytoplasmic Ca2+ in inactivated T cells was found to be increased. Since both the production of NO and the elevation of cytoplasmic Ca2+ are capable of inducing the clonal expansion of antigen-specific T 48 cells, it appears that histamine can negatively mediate this effect and contribute to immunosuppression [326].

Apart from its effect on T cells, there is increasing evidence to show that histamine has influences on the activation and maturation of DC with consequent impact on T cell polarisation [327]. Both the levels of HDC expression and histamine content were found to be increased during the differentiation of elutriated monocytes towards DC. In parallel with this finding, elevated surface markers of DC (CD80, CD86, CD40, CD45, and CD11c) are impaired by the inhibition of HDC along with the suppression of histamine binding [320]. Further, histamine has been shown to alter the repertoire of cytokines and chemokines in LPS-matured DC. In particular, the stimulation of H2R by histamine increases IL-10 and IL-8 production, whereas it suppresses IL-12 and IL-6 production in LPS-matured human DC. Thus, histamine is capable of polarising uncommitted maturing DC into the DC2 phenotype, which favours the differentiation of naive CD4+ T cells towards Th2 cells [328, 329]. Conversely, in the TNF-α-matured DC, histamine was instead found to increase IL-12 and IL-6 production, suggesting the involvement of different pathways [330]. In addition to the H2R, the H4R has also been -/- demonstrated to be involved in the polarisation of Th cells. DC isolated from H4R mice have a reduced production of IL-6, while their ability to induce Th2 differentiation is limited [324, 331].

Histamine and its receptors are implicated in a variety of pathogenesis. Evaluating transcriptional profiling of MS lesions showed that H1R expression is elevated compared to normal tissues [332]. Similarly, the susceptibility of EAE is associated with Bordetella pertussis-induced histamine sensitisation (Bphs), a gene later reported to be H1R [333]. In addition, H1R deficient mice exhibit significantly reduced EAE susceptibility, in which a suppression of IFN-γ and an increase of IL-4 is observed [333].

In line with the observation, H1R antagonists have been shown to maintain the clinical stability of MS patients, as well as reduce the severity of EAE [334, 335].

Whilst H1R antagonists are widely used in the treatment of allergy, H2R antagonists are effective in treating gastrointestinal disorders [336]. Interestingly, neither H1R nor H2R deficiency could rescue the development of arthritis in a mouse model, whereas HDC deficiency was shown to reduce the severity of arthritis [337]. These findings thus draw 49 the attention towards H4R. Genetic deletion or pharmacological blockade of the H4R has been observed to improve symptoms of inflammatory diseases such as airway inflammation [324, 338], pruritus [339], and asthma [340]. Additionally, in the model of zymosan-induced inflammation, H4R antagonists significantly reduced neutrophil infiltration. The effect is correlated with H4R-mediated down-regulation of leukotriene B4, which is a potent chemoattractant for neutrophils [341].

Besides, histamine signals through H4R have been shown to participate in regulating

Treg migration. During inflammatory demyelinating disease, the deficiency of H4R impairs the anti-inflammatory response due to a decreased frequency of CD25+Foxp3+ Treg and an increased proportion of Th17 cells in the CNS [342]. Likewise, it has been observed that H4R agonists suppress asthmatic inflammation by preferentially + + promoting the recruitment of CD25 Foxp3 Treg [343]. These H4R-responsive Treg exhibit potent suppressive activity on autologous T cells through an IL-10 dependent pathway. Collectively, the ability of HDC to modulate histamine catabolism can contribute to the regulation of inflammatory responses in various ways, by targeting distinct histamine receptors.

1.2.3.4 Phenylalanine metabolism by Interleukin-4 Induced Gene 1 (IL4I1)

Interleukin-4 Induced Gene 1 (IL4I1) has been demonstrated to possess L-amino acid oxidase (LAAO) properties. IL4I1 has preference for aromatic amino acid substrates as it catalyses the stereospecific deamination of the L-amino acid substrate, phenylalanine, to produce H2O2 and ammonia [344].

The expression of IL4I1 is restricted to lymphoid tissues, primarily in the lymph nodes and spleen [345, 346]. IL4I1 was first described in mouse and human B lymphocytes after IL-4 induction [345, 347]. Later, monocyte-derived dendritic cells and macrophages were identified that expressed IL4I1 when stimulated by pro- inflammatory mediators, such as IL-1β, TNF-α, and IFNs, and were considered to be the main producer of this enzyme [348].

IL4I1 exerts its function by inducing EAA depletion in the microenvironment and increasing the accumulation of the cytotoxic product, H2O2. The presence of H2O2 50 produced by IL4I1 suppresses the proliferation of target cells both in vitro and in vivo.

Mechanistically, H2O2 down-regulates the expression of CD3ζ in T cells, while the decrease in CD3ζ subsequently results in the suppression of the antigen-specific T-cell response [344, 349]. By inhibiting the CD8+ T cell response, IL4I1 was found to facilitate tumour escape from immune defence [349]. However, induced by TLR activation, IL4I1 also participates in an antibacterial effect by the induction of phenylalanine deprivation as well as the promotion of H2O2 secretion [350]. Also, the expression of IL4I1 in human Th17 cells has been shown to contribute to their self- regulation of expansion [351]. In a retinoid acid related orphan receptor (RORC) dependent manner, both un-stimulated and activated Th17 cells express high levels of IL4I1, which induces a decrease in CD3ζ and consequently impairs Th17 cell proliferation. So, the silencing of Il4I1 substantially up-regulates the proliferation of Th17 cells and their ability to produce IL-2, indicating another potential IL4I1- mediated pathway involved in the regulation of inflammation.

1.3 Role of MSC in haematopoiesis

HSC are capable of giving rise to all blood cell lineages throughout life. Three different states of HSC have been identified in the specialised microenvironment where they exist: 1) in a dormant state, where HSC are maintained in the G0 cell cycle phase and are considered to be metabolically inactive; 2) in a homeostatic state, where HSC occasionally enter the circulation but remain self-renewing; 3) in an injury-activated state, in which HSC are constantly cycling with high metabolism [352, 353]. Depending on the requirements of the system (whether it is under homeostasis or repair), the balance of these three states can be modified. To regulate this delicate balance, the key feature of niche is therefore associated with: 1) the long-term maintenance of HSC, as well as the regulation of their self-renewal and differentiation abilities; 2) homing capability, as HSC return to the bone marrow after tissue repair [352, 354].

A complex regulation is associated with HSC maintenance which relies on the interaction of HSC with the cellular component of the niche [355-357]. Currently, two niche models have been proposed, according to their localisation in the bone marrow (Figure 1.7). The endosteal niche is localised adjacent to the bone surfaces [353]. The endosteum, the cellular lining of bone surfaces separating the bone from the bone 51 marrow, is composed of osteoblasts and other stromal cells of mesenchymal origin. It has long been appreciated that the endosteum is a microenvironment which supports the dormancy of HSC, and so constituting a reservoir of HSC.

The vascular niche, which is associated with sinusoidal endothelium, is mainly composed of sinusoidal endothelial cells, perivascular reticular cells, and other cells of mesenchymal origin. It provides an intermediate niche for hosting dividing, self- renewing HSC that are ready to either differentiate into progeny or revert to dormancy, according to the needs of the organism. In order to fulfil different requirements, the endosteal and vascular niches are believed to cooperate in order to maintain an appropriate balance between the dormant and activated HSC. However, the question remains as to whether these two niches exist as separate entities physically and functionally, or whether they should be considered to be a ‘common’ niche controlled by both endosteal and perivascular stimuli [358].

Figure 1.7. Model of HSC niches in the bone marrow. In the endosteal niche, N-cadherin-expressing osteoblasts participate in HSC maintenance via producing angiopoietin-1 (Ang-1), bone morphogenetic protein 4 (BMP4), and Jagged-1. In addition, osteoblasts secret a range of cytokines supporting the expansion of haematopoietic progenitors, including granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-6 (IL-6), IL-1 and transforming growth factor- β (TGF-β). On the other hand, adipocytes have been shown to negatively regulate HSC maintenance. In the vascular niche, sinusoidal endothelial cells (SEC) can regulate haematopoiesis by producing angiocrine factors, such as fibroblast growth factor 2 (FGF2), BMP4, Ang1, and insulin-like growth factor-binding protein 2 (IGFBP2). Additionally, SEC mediate the mobilisation and homing of HSC via adhesion molecules, such as endothelial-cell (E)-selectin and vascular cell adhesion protein-1 (VCAM-1). Megakaryocytes (MK) are considered to maintain the quiescence of HSC by producing chemokine (C-X-C motif) ligand 4 (CXCL4), TGF-β, and Ang-1. Nestin+ MSC have been identified as a critical niche component as they express high levels of HSC maintenance factors, including CXCL12, stem cell factor (SCF), Ang-1, IL-7, osteopontin (OPN), and VCAM-1. Interestingly, CXCL12 abundant reticular (CAR) cells also produce high levels of SCF and CXCL12, while their localisation is highly overlapped with Nestin+ MSC.

1.3.1 The endosteal niche

Early study has demonstrated that the co-transplantation of osteoblasts with lineage negative (Lin-) BM cells enhanced engraftment and haematopoietic reconstitution in allogenic recipient mice [359]. Also, transplanted HSC were found to preferentially localise to blood vessels in endosteal regions [360]. The direct evidence showing that osteoblasts are fundamental in HSC regulation and maintenance came from two studies using transgenic mouse models.

In the first study, the administration of transgenic mice expressing constitutively active parathyroid hormone (PTH) or the PTH/PTH-related protein receptor (PPR) resulted in a simultaneous increase in the number of both osteoblasts and HSC in the BM [361]. In line with this observation, the addition of PTH to stromal cultures in vitro increased the number of osteoblastic cells and their ability to support ‘long-term culture-initiating cells’ (LTC-IC), which resembles HSC function in vivo in a linear manner. The study also indicated that treatment of recipient mice with PTH following BM transplantation markedly improved survival compared to mice treated with vehicle. In the second study, it was reported that the increase in osteoblast number is associated with a parallel increment in HSC [362], and in particular, a subset of osteoblasts expressing a high level of N-Cadherin function as a key component.

These cells were found to be located on the bone surface and were termed spindle- shaped N-cadherin+CD45-osteoblastic (SNO) cells. In BMP receptor type IA (BMPRIA) conditional knockout transgenic mice, an increase in the number of SNO cells was found to be correlated with an increase in the number of HSC. This was further supported by a study which led to the conclusion that an overall increase in the number and function of osteoblasts without a concomitant increase in N-cadherin+ cells is not sufficient to promote HSC quantity and function [363]. Nonetheless, several studies have encountered difficulties in detecting N-cadherin expression in HSC using gene expression profiling, flow cytometry, or β-galactosidase staining of N-cadherin gene trap mice [364, 365]. In addition, genetic deletion of Cdh2 (encoding N-cadherin) from HSC or from osteoblast lineage cells had no effect on HSC frequency, function, or haematopoiesis [366-368]. Together, these data challenged the notion of an N- cadherin+ osteoblastic niche. 54

Still, the importance of osteoblasts in HSC regulation remains viable. Accordingly, osteoblasts produce a whole range of cytokines, important for the expansion of haematopoietic progenitors while preserving their long-term culture-initiating activity in vitro, including G-CSF, GM-CSF, IL-6, IL-1 and TGF-β [369]. Additionally, osteoblasts express several cell-signalling molecules, including angiopoietin-1 (Ang- 1), bone morphogenetic protein 4 (BMP4), and Jagged-1, which are all involved in HSC self-renewal, survival and maintenance [354]. It was shown that Ang-1 expression in osteoblasts contributes to the interaction with Tie-2, a receptor tyrosine kinase expressed by HSC, which consequently promotes the quiescence of HSC [370]. Furthermore, the Tie-2/Ang1 interaction has been observed to activate β1-integrin and N-cadherin, which enhances the adhesion of HSC to the bone, thus protecting HSC from myelosuppressive stress. BMP4 has been recognised to be a critical component of HSC maintenance in the niche [371]. BMP4-deficient mice were found to obtain less c-Kit+ Sca-1+Lin- cells, in addition, BMP4-deficient mice recipients are associated with a microenvironmental defect that decreases the repopulating activity of WT HSC after transplantation. Jagged-1, as a Notch ligand, is capable of activating the Notch signalling pathway to enhance HSC self-renewal [372, 373]. However, a study showing that Notch-1 deficient HSC still efficiently reconstituted the BM chimera in the absence of Jagged-1 expressing BM stroma, led to the suggestion that Jagged-1-mediated Notch signalling is dispensable.[374].

1.3.2 The vascular niche

An intimate relationship between haematopoiesis and the vasculature is established during embryonic and fetal development, before the creation of BM cavities. The fact that HSC are capable of self-renewal and differentiation during this period further supports the existence of a vascular niche. Indeed, HSC have been observed to reside and to undergo differentiation in association with blood vessels in the placenta [375]. Furthermore, cell lines or purified primary endothelial cells generated from the yolk sac or the aorta-gonad-mesonephros (AGM) have been shown to support the maintenance and expansion of adult HSC in vitro [376, 377]. Although it is arguable that such association is displayed only during development and not in the adult, there is compelling evidence to suggest the existence of a perivascular niche.

BM sinusoidal endothelial cells (SEC) play a fundamental role in regulating haematopoiesis. Within the BM, vascular endothelial growth factor receptor 2 (VEGF2) expression distinguishes a continuous network of arterioles and SEC. The conditional deletion of VEGF2 in adult mice blocked the regeneration of SEC in irradiated animals, which consequently impairs the engraftment of HSPC and the replenishment of haematopoiesis [378]. Selective activation of Akt in the endothelial cells of adult mice improved the reconstitution of HSPC and accelerated haematopoietic recovery after myeloablation through the up-regulation of angiocrine factors, including fibroblast growth factor 2 (FGF2), BMP4, Ang1, and insulin-like growth factor-binding protein 2 (IGFBP2) [379]. In addition to their role for HSC maintenance, SEC contribute to the mobilisation and homing of HSPC via producing adhesion molecules, such as endothelial-cell (E)-selectin and VCAM-1 [380]. Further, SEC have been proposed to support the survival, long-term proliferation and differentiation of myeloid and megakaryocytic progenitors. For instance, SEC constitutively express CXCL12 (also called stromal-derived factor- 1, SDF-1) and FGF-4, which supports megakaryocyte (MK) maturation and thrombopoiesis [381].

Together with SEC, MK, reticular cells and mesenchymal progenitor cells are considered as critical components in perivascular niche. It has been demonstrated that ~20% of cells expressing HSC surface markers are located adjacent to MK in mice. Ablation of MK induced HSC cycling and resulted in increased HSC numbers, suggesting their role in maintaining HSC quiescence [382, 383]. This regulation can be mediated by several factors produced by MK, including CXCL4, TGF-β, and Ang-1 [384].

Other candidate cell types have been recognised in the perivascular niche by targeting CXCL12-CXCR4 signalling, which is essential for homing and maintenance of HSC. Immunohistochemical analysis has shown that high CXCL12 expression is expressed by a small population of VCAM-1+ reticular cells which remain in close contact with HSC [385]. A study has demonstrated that CD146-expressing human mesenchymal progenitors, localised in the BM sinusoids, were exclusively capable of establishing a haematopoietic microenvironment in developing heterotopic BM upon transplantation [33]. In addition to being localised near HSC, these cells express high levels of genes encoding CXCL12 and stem cell factor (SCF), suggesting their pivotal role in 56 haematopoietic maintenance [386]. The role of these cells will be better described in the following section.

When the homeostasis is disrupted by irradiation or oxidiative stress, HSC are mobilised from their quiescent state near the endosteum and migrate towards the vascular region near the sinusoidal endothelium [381]. Thus, it was proposed that the vascular niche promotes the proliferation and differentiation of actively cycling HSC, and their short-term nature. However, in one study early precursors of HSC were also found to be attached to the sinusoidal endothelium within the centre of BM [365], casting doubt on this notion. Collectively, the original view that osteoblasts are critical in regulating HSC maintenance and differentiation in the HSC niche has been revised, whereas the relevance of perivascular cells has been revealed.

1.3.3 MSC in the haematopoietic niche

For long time MSC have long been suggested to be involved in the regulation of haematopoietic progenitors, as mixed cultures derived from the adherent fraction of the BM stroma support the maintenance of HSC in vitro [387]. Not only MSC function as precursors of osteoblasts, also their ability to differentiate into adipocytes can regulate haematopoiesis. Adipocytes have been considered to be negative regulators of HSC. In genetically modified A-ZIP/F1 mice, which are incapable of forming adipocytes, or in mice treated with proliferator-activated receptor-γ inhibitor, which inhibits adipogenesis, BM engraftment after irradiation was enhanced compared to WT mice. The molecular mechanisms can be attributed to the secretion of neurophilin-1, lipocalin 2, adiponectin and TNF-α, each of which can impair haematopoietic proliferation [388]. Thus, MSC embedded in the BM niche are considered to modulate the regulation and maintenance of HSC in interconnected ways, but primarily via the collective action of soluble factors.

MSC produce several prominent growth factors, cytokines and chemokines [389, 390]. IL-6 was observed to work synergistically with IL-3 and SCF to promote the proliferation of human HSC [391]. Additionally, IL-3 and IL-6 function in combination with thrombopoietin (TPO), SCF, and Fms-like tyrosine kinase 3 ligand (FL) to support the ex vivo expansion and differentiation of HSC [392]. To maintain HSC in a quiescent 57 state, MSC produce early acting cytokines that preserve non-cycling HSC in the stem cell compartment [393]. Among those, the expression of leukemia inhibitory factor (LIF), SCF, FL and TGF-β1 has been demonstrated to promote the self-renewal of HSC rather than their differentiation. LIF has been demonstrated to support the survival and/or self-renewal of HSC in vivo, whereas a LIF deficient environment drives the differentiation of HSC [394]. SCF and FL are capable of enhancing the ex vivo expansion of human cord blood haematopoietic progenitors while remaining their self- renewal capacities [395]. In addition, FL favours the amplification of LTC-IC population in CD34+CD38- human BM cells [396]. TGF-β1 has been documented as a potent growth inhibitor that maintains CD34+ HSC in quiescence by down-modulating differentiation-associated cytokine receptors, such as granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage colony-stimulating factor (GM-CSF), IL-1, and IL-3 [397]. Mechanistically, gp-130, a common cytokine receptor subunit, is considered to mediate the effect of these cytokines and utilise gp-130-linked signal transduction pathways.

The notion that MSC are part of the HSC niche is further supported by a finding, in which a nestin-expressing MSC subpopulation residing on the walls of BM sinusoids in close proximity to HSC was identified [398]. These cells are spatially associated with adrenergic nerve fibres of the sympathetic nervous system (SNS), which, by regulating circadian rhythm, controls oscillation in circulating HSC numbers [398, 399]. Nestin+ MSC express high levels of HSC maintenance factors, including CXCL12, SCF, Ang- 1, IL-7, VCAM1, and osteopontin (OPN). These factors are also known to induce osteoblastic differentiation of MSC. Nestin+ cell depletion resulted in an increased mobilisation of HSC to the periphery in vivo and decreased the homing of HSC after transplantation. In addition, CXCL12 production by nestin+ MSC plays an important role in HSC niche formation during BM development [400]. Conditional depletion of Cxcl12 in MSC led to an approximately 30% reduction of HSPC in neonatal BM measured by long-term competitive repopulation assays.

Sugiyama et al. demonstrated that reticular cells with high levels of CXCL12 expression, termed CXCL12 abundant reticular (CAR) cells, are in close contact with the majority of HSC regardless of their endosteal or perivascular localisation [385]. Deletion of CXCR4, the receptor for CXCL12 in mice, led to a reduction in HSC 58 numbers [385]. Selective short-term ablation of CAR cells inhibited HSC proliferation by reducing the production of SCF and CXCL12, accompanied by an increase in HSC quiescence and myeloid differentiation [401]. Although the exact composition of CAR cells remains unknown, it has been demonstrated that the majority of CAR cells express adipogenic and osteogenic genes. The expression of myxovirus resistance-1 (Mx1), an osteogenic marker, has been observed in CAR cells, but also in stromal cells enriched for mesenchymal stem and progenitor cells (MSPC) activity [402]. CAR cells have also been identified to express leptin receptor (LepR). LepR-expressing CAR cells share similar phenotypes with LepR-expressing MSC, while both cell types produce high levels of SCF and CXCL12 [403, 404]. The produced CXCL12 by LepR+ cells is required to retain HSC and colony-forming progenitor in the BM [405]. Interestingly, Lepr-expressing cells were found to largely overlap with nestin+ stromal cells using immunofluorescence imaging [406]. These findings suggest that CAR cells are adipo- osteogenic progenitors that not only contribute to adipocyte and bone generation, but also required for HSC maintenance and functions.

The question remains as to whether CAR cells are the same as nestin+ MSC, or whether these are two highly overlapping CXCL12-expressing cell populations; two theories exist. Firstly, nestin+ MSC might represent a primitive subtype of CAR cells as they exist in much smaller numbers compared with CAR cells, they display colony-forming unit fibroblast activity, they harbour high self-renewal activity both in vitro and in vivo, and they retain multilineage differentiation capacity.[407]. Secondly, CAR cells might represent a more committed precursor, modulating haematopoietic progenitors rather than HSC. In fact, depletion of CAR cell was found to be lethal within 5 days, and affected numerous types of haematopoietic cells, whereas depletion of nestin+ cells seems to affect HSC only [408].

1.3.4 Role of MSC in HSC transplantation

Due to their ability to regulate haematopoiesis and to differentiate into niche cells, MSC have been investigated for their potential role in supporting the engraftment of HSC following transplantation. Indeed, co-transplantation of MSC and cord blood or mobilised peripheral blood CD34+ cells accelerates and increases the engraftment when compared to the transplantation of CD34+ cells alone [409, 410]. This enhancement was 59 further amplified by transplantation of GM-CSF and SCF-transfected MSC, suggesting that these HSC-promoting factors are associated with engraftment, albeit with the underlying mechanism remaining to be documented [411]. Nevertheless, the effect of MSC on promoting HSC engraftment is MHC-restricted, as the formation of cobblestone colonies of HSC with MHC-mismatched stromal cells was reduced in comparison with MHC-matched stromal cells [412]. The first clinical trial demonstrated an accelerated haematopoietic recovery after co-transplantation of autologous MSC and HSC in advanced breast cancer patients after the receipt of high- dose chemotherapy [413]. However, in other allogeneic HSC transplantations, even when donor and recipient are HLA identical siblings, beneficial effects of MSC were not observed [414].

GvHD is one of the major challenges of allogeneic HSC transplantation. It results from the generation of cytotoxic effect of donor T cells attacking host tissues. The infusion of MSC reduces the incidence of GvHD by controlling the homeostasis of T subset [415, 416]. These data prompted investigations into clinical transplantation, with the clinical efficacy appearing to be inconsistent [414, 417]. In a phase II clinical trial, 39 of 55 patients with steroid resistant grade IV acute GvHD responded to MSC treatment, showing higher survival and lower transplantation-related mortality. In contrast, a phase III study in steroid refractory acute GvHD failed to show a difference between the group receiving MSC and the controls. This discrepancy was later explained by the findings from animal studies. Whilst in an murine acute GvHD (aGvHD) model, a single infusion of MSC at the time of HSC transplantation was not effective in preventing aGvHD [418], this could be mitigated by administrating multiple doses of MSC at weekly intervals subsequent to HSC transplantation [419]. In addition, MSC was reported to treat aGvHD when administered at the peak of IFN-γ production following donor T cell recognition of antigen, suggesting the importance of IFN-γ to initiate MSC efficacy [420]. Thus, these findings strengthen the concept of MSC ‘licensing’, indicating the necessity of appropriate inflammatory environment which enables MSC to exhibit their immunomodulatory properties.

1.3.5 Contribution of MSC to haematopoietic malignancies

As an essential component of the BM niche, MSC are also instrumental at creating a favourable microenvironment for tumour development. It is believed that a leukaemia stem cells (LSC) is a fundamental component of the malignant process because of its ability to self-renew. This activity is mainly mediated by the constitutive activation of signalling factors, including NF-κB, Akt and Wnt/β-Catenin [421-423].

MSC have been shown to regulate the dormancy of LSC [424-426]. The cell cycle status of LSC critically influences the relative drug sensitivity and resistance in haematological malignancies, such as acute myelogenous leukaemia (AML) and CML. It has been shown that quiescent cells isolated from primary AML specimens displayed potent repopulating capability after transplantation into immune-deficient mice [427]. Likewise, primary CML specimens exhibit a primarily quiescent LSC population; in addition they appear to reside near to the endosteal surface upon transplantation [428, 429]. This explains the resistant and relapse phenomenon observed in patients where only proliferative leukaemia cells are eliminated by cell cycle-dependent chemotherapeutic drugs, and the dormant LSC remain unaffected. As a consequence, inducing the cell cycle progression of LSC by interrupting their interactions with MSC might potentially improve the efficacy of chemotherapy.

The interactions between LSC and MSC involve both cell contact interactions and soluble factors. For instance, the levels of CXCR4 are highly elevated in different types of leukaemia, including acute lymphoblastic leukaemia (ALL) and AML, and CXCR4 expression is correlated with poor outcomes [430-432]. Concomitantly, CXCL12 is overexpressed in MSC isolated from patients with MDS, and represents an important determinant of leukemic quiescence [433]. CXCR4/CXCL12 signalling is associated with the up-regulation of pro-survival signals and quiescence, both contributing to resistance to chemotherapy. Blocking CXCR4 in vitro using polypeptide antagonists resulted in decreased chemotaxis toward CXCL12-expressing cells and an inactivation of pro-survival Akt signalling. As a result, disruption of CXCR4/CXCL12 interaction enhanced the susceptibility of AML cells to chemotherapy-induced apoptosis [434]. In line with these observations, the administration of CXCR4 antagonists to NOD/SCID mice, which had been previously engrafted with primary AML cells, led to a dramatic 61 decrease in the levels of AML cells in the BM, blood, and spleen. Following the same treatment, the levels of normal human progenitors engrafted into NOD/SCID mice were not affected [435]. CXCL12-expressing MSC have been shown to protect CML cells from apoptosis induced by imatinib, a tyrosine kinase inhibitor. In fact, a CXCR4 antagonist was shown to restore the sensitivity of CML cells to imatinib [436].

TGF-β1 is another regulator that has been demonstrated to promote the quiescence and survival of leukemic cells. MSC were shown to support the survival of primary leukaemia cells accompanied with cell cycle arrest via activation of the PI3K/Akt pathway and subsequent up-regulation of C/EBPβ (a member of the leucine zipper family of transcriptional regulators with fundamental roles in cell proliferation) expression, differentiation and senescence [437]. Instead, a TGF-β1 inhibitor of small molecular size (LY2109761) reduced the anti-apoptotic effects of MSC. Hypoxia and interactions with MSC are considered to result in a microenvironment which promotes leukaemia cell survival and leads to chemoresistance. Exogenous TGF-β1 has been documented to maintain AML cells in a quiescent state, which was further facilitated in the presence of MSC. In turn, blocking TGF-β1 with neutralising antibody (1D11) abrogates such an effect and enhanced the apoptosis of AML cells under hypoxic and normoxic conditions [438]. In combination with a CXCR4 antagonist, administration of 1D11 produced a remarkable effect enhancing the efficacy of chemotherapy against AML cells and prolonging the survival of mice under hypoxic conditions.

1.4 Hypothesis and aims

The hypothesis guiding this research is that, by regulating EAA deprivation, MSC regulate the proliferation of immune cells, HSC and tumour cells. A set of specific aims was generated to examine this hypothesis:

1. The profile of EAA consuming enzyme in MSC-mediated immunosuppression under inflammation conditions. 2. The underlying mechanisms responsible for the increased enzyme production by MSC. 3. The effect of EAA deprivation on maintaining the quiescence of HSPC by MSC. 4. The role of candidate molecules in MSC-mediated anti-apoptotic effect on tumour cells.

2 MATERIALS AND METHODS

2.1 Chemicals and Reagents

N-[[3-(aminomethyl)phenyl]methyl]-ethanimidamide, dihydrochloride (1400W), S-(2- boronoethyl)-L-cysteine (BEC) were purchased from Cayman Chemicals (USA). α- methyl-DL-histidine dihydrochloride, etoposide were purchased from SIGMA. 3-[[4- (1,1-dimethylethyl)phenyl]sulfonyl]-2E-propenenitrile (BAY11-7085), [5-(p- Fluorophenyl)-2-ureido]thiophene-3-carboxamide (TPCA-1), N-(6-Chloro-7-methoxy -9H-pyrido[3,4-b]indol-8-yl)-2-methyl-3-pyridinecarboxamide dihydrochloride (ML- 120B), and PF-026 was kindly provided by Dr. Mark Birrell (Imperial College). Recombinant mouse IL-1β, IL-4, IL-13, TNF-α, and IFN-γ were obtained from Peprotech (London, UK). Poly (I:C) and LPS were purchased from Sigma.

2.2 Animals

C57BL/6 and BALB/c mice were purchased from Olac, Bistar UK. Nos2-/- mice in C57BL/6 background were kindly provided by Prof. Vincenzo Bronte (Verona University Hospital and Department of Pathology, Immunology Section, Verona). Tnfr1/r2-/- mice in C57BL/6 background were kindly provided by Dr. Jorge Caamano (Medical Research Council (MRC) Centre for Immune Regulation, Institute for Biomedical Research, Birmingham Medical School, Birmingham). Myd88-/- mice in C57BL/6 background were kindly provided by Dr. Neil Rogers (Immunobiology laboratory, London research institute, Cancer research UK). Mice were maintained at CBS Unit, Imperial College London, in accordance with Home Office guidelines (Animals [Scientific Procedures] Act 1986). BM and spleen cells were obtained from 8-12 week old mice.

2.3 Cells and cell culture

2.3.1 Generation of MSC

MSC were generated from bone marrow of 8 to 12 weeks old mice. Femurs and tibias of the mice were directly crushed and plated in Mesencult. After 72 hours non-adherent cells were removed and adherent cells were maintained with medium replenishment every 3 to 4 days. After 8 to 10 weeks, a homogeneous cell population was obtained. The identity of MSC was confirmed by immunophenotypic criteria based on the expression of PDGFRα and Sca-1, and the absence of haematopoietic marker CD45 (eBioscience).

2.3.2 Preparation of splenocytes

Splenocytes were harvested from freshly euthanized C57BL/6 mice. Single-cell suspensions were generated by passing spleens through a cell strainer (70μm, BD Bioscience, UK), washed with complete RPMI, and centrifuged at 1500 rpm for 6 minutes. Following red blood cell lysis with ammonium chloride (Lonza), cells were washed and collected by centrifugation, refiltered, enumerated, and assayed for viability with trypan blue (SIGMA). Activated splenocyte supernatant was harvested from 48-hour cultures of splenocytes (2x106/ml) activated by 3μg/ml concavalin A (Con A), then the supernatant was filtered and frozen at -80°C.

2.3.3 Cell culture

All cell culture procedures were performed under sterile conditions in a flow cabinet.

Cell cultures were incubated in 5% CO2, at 37ºC, in a humidified incubator.

MSC were cultured in Mesencult medium with MSC stimulatory supplement (StemCell Technologies, Canada). Splenocytes were cultured in RPMI 1640+GlutaMAXTM medium supplemented with 10% heat inactivated foetal bovine serum (FBS) (Biosera, South America), 100U/ml penicillin G, 100U/ml streptomycin sulphate, and 25µM β-

65 mercaptoethanol (β-ME) (Invitrogen). HSPC were cultured in in StemSpan® SFEM medium with or without 50ng/ml SCF, 50ng/ml FLT-3L, 10ng/ml IL-3 (Peprotech). Murine erythroleukemia (MEL) cell line, which is erythroid progenitor cells derived from the spleens of susceptible mice infected with the Friend virus complex, was a gift from Dr. Vincent Kao (Department of Oncology Research, King’s College London, London). MEL cells were cultured in RPMI 1640+GlutaMAXTM medium supplemented with 10% heat inactivated FBS (Biosera, South America), 100U/ml penicillin G, and 100U/ml streptomycin sulphate (Invitrogen). EL4 is a T lymphoma cell line established from the tumour induced by 9,10-dimethyl-1,2-benzanthracene in C57BL/6 mice. EL4 cells were cultured in DMEM medium supplement with 10% heat inactivated FBS (Biosera, South America), 2mM glutamine, 100U/ml penicillin G, 100U/ml streptomycin sulphate, 10mM Hepes, and 25µM β-ME (Invitrogen).

2.3.4 Co-culture

MSC and splenocytes

MSC were plated at 1.25x105 cells/well in 24 well plates (Costar), and kept at 37°C,

5% CO2 and 95% humidity in a cell incubator. When the MSC were attached to the plastic, 5x106 naïve or Con A-stimulated splenocytes were added in a 0.4μm transwell system (Corning, Costar).

MSC and HSPC

MSC were plated at 1x105 cells/well in 24 well plates (Costar), and kept at 37°C, 5%

CO2 and 95% humidity in a cell incubator. HSPC were isolated from WT BM cells by EasySep™ Mouse Hematopoietic Progenitor Cell Enrichment Kit. Briefly, BM cells were stained with biotinylated antibodies against lineage antigens (CD5, CD11b, CD19, CD45R, Ly-6G/C, Ter119, CD71), followed by removal of lineage-positive cells using a magnetic bead protocol. When the MSC were attached to the plastic, 4x105 HSPC were co-cultured with MSC for 5 days. Cells were cultured in in StemSpan® SFEM medium containing 50ng/ml SCF, 50ng/ml FLT-3L, 10ng/ml IL-3 (Peprotech). After collection, cells were firstly labelled with CD45-APC (eBioscience) to distinguish HSPC. In the experiments that HSPC were stimulated with 50ng/ml G-CSF (Peprotech), 66

HSPC were co-cultured with MSC in the StemSpan® SFEM medium without cytokine supplement.

MSC and EL4 cells

MSC were plated at 1.25x105 cells/well in 24 well plates (Costar), and kept at 37°C,

5% CO2 and 95% humidity in a cell incubator. When the MSC were attached to the plastic, 0.5x106 EL4 were plated into wells. 1μM or 15μM of etoposide (Sigma) was added to cell culture to induce cell death.

2.4 PCR

Genomic DNA was isolated from WT MSC or Myd88-/- MSC using the DNA Blood Midi Kit (Qiagen) according to manufacturer’s instructions. DNA concentration was measured using a NanoDrop 1000 spectrometer (Thermo Fisher Scientific, USA). In brief, PCR reaction conditions were as follows: 10 µl of DNA was added to 40 µl of a reaction mix containing 1.25l of forward and reverse primers (20M), 1l of dNTP mix (10 mM dATP, dCTP, dGTP, dTTP), 0.5µl of Taq polymerase (Thermo Fisher

Scientific), 5l of PCR reaction buffer (750 mM Tris-HCl, 200 mM (NH4)2SO4, 0.1%

Tween20, pH 8.8), 3l of MgCl2 (25mM), and 28l of ddH2O. PCR was performed using the following cycles: 94°C–5 minutes; 36 cycles of: 94°C–1 minutes, 52°C–2 min, 72°C–2 min; 72°C–10 min, hold at 4°C on a DNA Thermal Cycler (PerkinElmer Cetus, Warrington, United Kingdom). The following primer sequences were used: WT MyD88 gene; forward, 5′–ATTGCCAGCGAGCTAATTGAG-3’, and reverse, 5’- AGTCCTTCTTCATCGCCTTGT-3’, KO allele MyD88 gene; forward, 5’- ATTGCCAGCGAGCTAATTGAG-3’, and reverse, 5’- ATCGCCTTCTATCGCCTTCTTGACGAG-3’. The PCR products (10 µl) were resolved by electrophoresis in a 1.5% TAE agarose gel containing 1g/ml ethidium bromide. The gel was photographed under ultraviolet light.

2.5 Real-time PCR

MSC were lysed with RLT Plus buffer (Qiagen, Valencia, CA) supplemented with 1% β-ME (Sigma). Lysed cells were homogenized by passage through QIAshredder columns (Qiagen). DNA removal and total RNA isolation was performed with the RNeasy Plus Mini Kit (Qiagen) according to manufacturer’s instructions. RNA concentration was measured using a NanoDrop 1000 spectrometer (Thermo Fisher Scientific, USA). RNA samples were diluted to 25 ng/μl in diethyl pyrocarbonate- treated water. RT-PCR reactions were performed on an ABI Prism 7900HT instrument (Applied Biosystems, Foster City, CA) using the Taqman RNA-to-CT 1-Step Kit (Applied Biosystems), following the manufacturer’s instructions and with the following cycle conditions: 48°C–15 minutes, 95°C–10 minutes; 40 cycles of: 95°C– 15 seconds, 60°C–1 minutes. Primer/probe mixes were Mm00475988_m1 (Arg1), Mm00477592_m1 (Arg2), Mm00456104_m1 (Hdc), Mm00446968_m1 (Hprt1), Mm00492586_m1 (Ido), and Mm00515786_m1 (Il4i1), Mm00435175_m1 (Nos1), Mm00440485_m1 (Nos2) (Applied Biosystems). Real-time PCR was monitored and analyzed with Sequence Detection System version 2.0 (Applied Biosystems) Reverse transcriptase negative controls were performed in parallel and showed no amplification. All samples were run in triplicate and data were analyzed using the delta delta CT (ΔΔCT) relative quantification method, with each sample normalised to the Hprt1 housekeeping gene to correct for differences in RNA quantity.

2.6 Western blot

MSC were plated at 1x106 cells/well in 6-well plates (Costar), and kept at 37°C, 5%

CO2 and 95% humidity in a cell incubator. When the MSC were attached to the plastic, the medium was replaced by supernatant obtained from Con A-activated splenocytes or by medium containing cytokines as stated. After 24 hours, the cells were collected and resuspended in homogenization buffer of composition: 50mM Tris (BDH), 150mM NaCl (Sigma), 5mM EDTA (Sigma), 1% Triton X-100 (v/v) (BDH) and protease inhibitor mixture (Complete®, Roche Applied Science). For p65 nuclear translocation, nuclear and cytoplasmic fractionation was conducted using the NE-PER Nuclear and Cytoplasmic Extraction Reagents kit (Thermo Fisher Scientific) according to the

68 manufacturer’s protocol. Samples were centrifuged, and the supernatant was collected and assayed for total protein concentrations using the Bradford method (Bio-Rad) and analysed with a BioPhotometer Plus photometer (Eppendorf) at wavelength 550-650nm. Equal amounts of proteins were resuspended in 2x laemmli buffer (Sigma). Samples were then heated at 95°C for 5 min, allowed to cool at ambient temperature and loaded onto SDS-polyacrylamide gels (Bio-Rad). Electrophoresis was run at 8-16 mA. Proteins (25μg) were transferred onto a nitrocellulose membrane (Schleicher and Schnell, Hamburg, Germany) in a Trans-Blot® SD semi-dry transfer cell (Bio-Rad) using a current of 15V for 30-45 min. The blots were blocked with 5% (w/v) nonfat milk (Bio-Rad) in PBS buffer containing Tween 20 (0.05% w/v, NBS Biologicals) and incubated overnight with rabbit anti-NOS2 polyclonal antibody (ab15323) antibody (1:200, Abcam), or with rabbit anti-p65 (D14E12) XP® antibody (1:1000, Cell signaling). Blots were incubated with ECL Rabbit IgG, HRP-linked antibody (1:1000, Amersham), and immunoreactivity was visualized using an ECL detection system (Amersham).

2.7 Confocal

WT MSC or Nos2-/- MSC BM were grown on coverslips for 24 hours, untreated or treated with TNF-α (20ng/ml) and IFN-γ (20ng/ml). The cells were rinsed with PBS, fixed in 4% paraformaldehyde (PFA) overnight at 4°C, and airdried. The coverslips were incubated overnight with a solution containing anti-α-tubulin (1:250, Abcam) and anti-iNOS (1:100, Abcam) in PBS containing 1% BSA. The coverslips were washed three times with PBS, and then incubated with a solution containing Dylight 594- conjugated anti-Ig directed to the native species of the primary antibody (1:200) in PBS containing 1% BSA. This secondary incubation was done for 1 hour at room temperature. The coverslips were washed three times in PBS and mounted to the slides in Vectashield mounting medium containing 4,6-diamidino-2-phenylindole (DAPI) (Vector Laboratories, Inc., Burlingame, CA). The slides were then sealed with nail polish and stored at 4°C. Confocal images were obtained using a Leica TCS-SP laser- scanning confocal microscope (Leica, Heidelberg, Germany).

2.8 Immunosuppressive assay

Naïve or Con A-stimulated splenocytes (5x105 per well) were cultured with or without the pre-adherent MSC in a 96 well flat bottom plate (Costar). Inhibitors of EAA consuming enzymes were added as specified and cells were kept at 37°C, 5% CO2 and 95% humidity in a cell incubator. Cell proliferation was measured by incorporation of tritiated thymidine ([3H] thymidine). Cells were pulsed with 0.5μCi/well of [3H] thymidine (Amersham, UK) during the last 18 hours of culture. Cells were then harvested onto filters (Wallac Perkin Elmer, USA) using a 96-well cell harvester. Scintillation fluid (Wallac) was added to the filter, and the radioactivity was detected by the beta-counter (Wallac). The results of the incorporated thymidine were expressed as relative inhibition of the culture calculated by the following equation: inhibition= 1- (proliferation count of the sample/ mean proliferation count of the positive control).

2.9 NO quantification

- - To determine total NO production by MSC, the level of nitrite (NO2 ) and nitrate (NO3 ) were measured using Nitrate/nitrite colorimetric assay kit (Cayman Chemical). MSC were plated at 1.25x105 cells/well in 24-well plates (Costar). When the MSC were attached to the plastic, the medium of MSC was replaced by supernatant obtained from Con A-activated splenocytes. Samples were collected after 24 hours of culture. The assay was performed as described by the manufacturer. Briefly, 80μl of the sample was mixed with 10μl cofactor and 10μl nitrate reductase. After two hours of incubation at room temperature to convert nitrate into nitrite, total nitrite in the sample was measured at 540nm absorbance by reaction with Griess reagent (sulfanilamide and naphthalene– ethylene diamine dihydrochloride).

2.10 Cell viability

Cells were firstly labelled with CD45-FITC to distinguish EL4 cells. Then, the cell apoptosis was assessed by Annexin V / 7-Amino-Actinomycin (7-AAD) staining (PE Annexin V Apoptosis Detection kit, BD Biosciences) and FACS analysis, following the manufacturer’s protocol. Briefly, cells were washed twice with cold PBS and then

70 resuspended in Binding Buffer (10 mM Hepes/NaOH (pH 7.4), 0.14 M NaCl, 2.5 mM 6 CaCl2.) at a concentration of 1 x 10 cells/ml. 5 μl of PE Annexin V and 5 μl of 7-AAD were added to 1 x 105 cells, followed by 15 min incubation at room temperature in the dark. 400 μL of binding buffer were added to each sample that was analyzed by flow cytometry within 1 hr.

2.11 Quantification of HY-primed T cells stimulated in vivo

Female C57BL/6 mice were immunised intra-peritoneally (i.p.) with 15x106 male splenocytes at day 0 and day 7. Mice were sacrificed after 14 days. The splenocytes were harvested and seeded in flat bottom 96-well plates at a final concentration of 0.5x106 cells per well as responder cells. Irradiated female splenocytes (3000 rads) were incubated with 0.1, 1, 10, and 100 μg/ml H-Y peptide, HYDb Uty or HYDb Smcy (ProImmune), for 1 hour and washed 3 times with PBS. Then, the cells were plated at 2x106, 1x106, 0.5x106 and 0.25 x106 cells per well as stimulator cells, to make 1:4, 1:2, 1:1, and 1:0.5 ratio of responder to stimulator cells. The cells were cultured for 72 hours, and [3H] thymidine at 1 μCi/ml was added for the last 18 hours. The proliferation was assessed by [3H] thymidine uptake as described.

When applying MSC to the same model, female C57BL/6 mice were immunised i.p. with 15x106 male splenocytes at day 0, day 7 and day 14. MSC (15x106) were injected i.p. at day 7 and on day 14. Mice were sacrificed after 21 days and the splenocytes were harvested and plated at 0.5x106 cells per well. Irradiated female splenocytes (3000 rads) were incubated with 0.1, 1, 10, and 100 μg/ml H-Y peptide, HYDb Uty or HYDb Smcy, for 1 hour and washed 3 times with PBS. Then, the cells were plated at1x106 cells per well. In parallel, irradiated male splenocytes (3000 rads) were plated at 2x106, 1x106, 0.5x106 and 0.25 x106 cells per well as stimulator cells. The cells were cultured for 72 hours, and [3H] thymidine at 1 μCi/ml was added for the last 18 hours. The proliferation was assessed by [3H] thymidine uptake as described.

2.12 Thioglycollate (TG)-induced peritonitis

Female C57BL/6 mice were injected i.p. with 1 ml of Brewer’s thioglycollate broth (Sigma) to induce a sterile peritonitis or with PBS as a control. 30 minutes after the first injection, 2x106 MSC were injected through the same route. After 3 days, the mice were sacrificed. Peritoneal exudate cells were collected by lavage with 10ml of ice-cold, sterile PBS. After harvesting, peritoneal cell numbers were determined by haemocytometer counts. Brewer’s Thioglycollate broth (4%) was prepared, autoclaved, and aged in the dark for at least 1 month prior to use.

2.13 FACS analysis

Nonspecific Fc receptor binding was blocked with PBS supplemented with 1% FBS, and mouse serum IgG before incubation with the respective monoclonal antibody (mAb) at 4°C for 20 minutes. After incubation the cells were washed with PBS and analysed by flow cytometry. Samples were acquired with a BD FACSCalibur or with a BD LSRII (BD Biosciences) and data was subsequently analysed using FlowJo software (Oregon, USA). A list of antibody used is represented in Table 2.1.

For cell-cycle analysis, cells were incubated for 45 minutes at 37°C with 1μg/ml Hoechst 33342 (Sigma). Then, pyronin Y (Sigma) was added at 0.5μg/ml, and the cells were incubated for another 15 minutes at 37°C. The cells were immediately acquired using a LSR II flow cytometer (BD Bioscience) and data was subsequently analysed using FlowJo software (Oregon, USA).

2.14 Statistical analysis

Data were analysed using unpaired two-tailed Student’s t test with confidence interval of 95%, and expressed as mean ± standard deviation (SD).

Antibody Clone Fluorochrome Manufacturer CCR2 475301 APC R&D CD11b M1/70 PerCP/Cy5.5 Biolegend CD140a (PDGFRα) APA5 APC eBioscience CD45 30-F11 APC eBioscience CD45 30-F11 FITC eBioscience F4/80 CI:A31 Alexa Fluor 700 AbDSerotec Ly-6C AL-21 FITC BD Biosciences Ly-6G 1A8 PE BD Biosciences Ly-6G/Ly-6C (Gr-1) RB6-8C5 APC Biolegend Sca-1 (Ly-6A/E) D7 FITC eBioscience

Table 2.1. List of monoclonal antibodies for FACS analysis

3 THE ROLE OF EAA METABOLISM IN MSC- MEDIATED IMMUNOSUPPRESSION

3.1 Introduction

MSC are present in virtually every organ and, together with macrophages, constitute the tissue niche that regulates inflammation and tissue homeostasis.

An emerging body of data demonstrates the effects of MSC on inhibiting the proliferation and function of immune cells including T, B, NK cells, as well as DC. Although the immunomodulatory properties of MSC are well established, the actual mechanisms by which MSC exert their functions remain to be defined. Several studies have been aimed at elucidating the mechanisms of MSC-mediated immunosuppression, and have indicated a role for TGF-β1, HGF, PGE2, IL-10, and HLA-G [181, 439-442]. However, blocking each of these factors alone does not completely restore the proliferation of activated immune cells, indicating that multiple factors and/or unknown factors are involved. EAA metabolism represents a potential key pathway. Regulating the exhaustion of EAA and inducing cell-cycle arrest is an important strategy for controlling the immune response and has been clearly demonstrated in subsets of tolerogenic DC. In tolerogenic DC, EAA consuming enzymes mediate the exhaustion of EAA and, therefore, inhibit T lymphocyte proliferation. On the basis of these notions, I wanted to explore the profile of EAA consuming enzymes, the activation of which may contribute to MSC-mediated immunosuppression.

In addition, the immunosuppressive activity of MSC is not constitutive, but requires a licensing signal. As MSC migrate to the injured sites where inflammation occurs, they can respond to different ‘danger’ signals. It is of note that the microenvironment that MSC are exposed to is crucial to their immune modulating activities. This feature of MSC modulation is accomplished by MSC expression of various cytokine receptors 74 and TLR, which polarise MSC into distinct phenotypes upon stimulation. A paradigm has been suggested in which MSC can be polarised towards pro- or anti- inflammatory phenotypes depending on the receptor stimulated. Based on these observations, I aimed to determine the impact of the inflammatory environment on the MSC expression of EAA consuming enzymes.

3.2 Results

3.2.1 Characterisation of isolated MSC

My first approach was to standardise a procedure for MSC generation and expansion in order to make the experiments comparable throughout the project. Cultures of BM derived MSC were established and evaluated for the expression of PDGFRα, Sca-1, c- kit, and the stromal marker, CD90, by FACS analysis at different intervals. The presence of contaminating haematopoietic cells was excluded by CD45 expression. A representative result from the characterisation of the immunophenotype of MSC is illustrated in Figure 3.1A. MSC were cultured for 8 weeks and underwent 5 passages. More than 96.6% of cells were negative for CD45, and the proportion of PDGFRα+/Sca-1+ cells within this population was 79.5%. The expression of the marker c-Kit was not detectable in BM derived MSC, whereas 95.6% cells were positive for CD90.

In order to assess the immunosuppressive ability of BM-derived MSC generated as described, their effect on suppressing CD3-induced T-cell proliferation was evaluated by [3H] thymidine uptake following 3 days of culture. Escalating concentrations of MSC were cultured with splenocytes induced by 3µg/ml Con A as mitogen or by 30µl/ml anti-CD3/CD28 beads (one bead per cell) (Figure 3.1B). Inhibition of T-cell proliferation by MSC occurred in a dose-dependent manner. MSC suppressed the proliferation of splenocytes induced by Con A in a similar fashion as the MSC-induced suppression observed in the presence of anti-CD3/CD28 beads (Figure 3.1B). The proliferation of Con A-activated splenocytes was completely abolished when MSC were present in a ratio of up to 1:160 , whilst the suppression of anti-CD3/CD28 beads stimulated splenocytes was partially reversed at a 1:160 ratio (90.95%±3.01%) (Figure 3.1B). The results demonstrate that the potent suppressive effect of MSC on T cells was acquired by stimulation triggered by either Con A or anti-CD3/CD28 beads.

Figure 3.1. MSC exhibit dose-dependent anti-proliferative activity. Splenocytes were plated at 0.5x106 cells/well and stimulated with (A) 3µg/ml of Con A or (B) 30µl/ml anti-CD3/CD28 beads in the presence or absence (controls) of C57BL/6 MSC at the ratio indicated for 3 days. 0.5 μCi/well of [3H] thymidine was added for the 18 hours prior to the harvesting of cells. One representative experiment out of 3 independent experiments is shown. Proliferation was determined as a measure of [3H] thymidine uptake, and results are expressed as percentage of inhibition, obtained using the formula: {1-

[(MSCSplConA)-(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD of cultures in triplicates.

3.2.2 MSC express a selective pattern of EAA consuming enzymes in response to activated splenocytes

Since the anti-proliferative activity of MSC is highly dependent on the inflammatory microenvironment, I investigated the expression of candidate enzymes in a system in which MSC were exposed to naïve or previously Con A-activated splenocytes (Figure 3.2) in a transwell system. The use of this system allowed me to rule out contaminating splenocytes in the MSC sample analysed [159, 165]. After 24 and 48 hours in culture, MSC were harvested from the bottom well and evaluated for the expression of EAA consuming enzymes including Arg1, Arg2, Nos1, Nos2, Ido, Hdc, and Il4i1 by quantitative RT-PCR. Whilst the incubation with naïve splenocytes did not induce the upregulation of any of the enzymes tested compared to control MSC, the MSC cultivated in the presence of activated splenocytes exhibited a defined profile.

It was observed that activated, but not resting splenocytes induced the transcription of 5 enzymes (Arg1, Arg2, Nos2, Hdc, and Il4i1) at different levels and with different kinetics. A significant increase (mean 6-fold) in Arg1 (6.23±1.89 versus MSC alone, p=0.0083) could be detected only after 48 hours. Arg2 expression was up-regulated (12-fold) after 24 hours but increased further after 48 hours (49.56±8.28 versus MSC alone, p=0.0016). The amino acid L-arginine is not only a substrate for ARG, but it is also catalysed by NOS. Although Nos1 expression was unaffected by the presence of resting or activated splenocytes, a marked Nos2 upregulation was observed at levels far superior to both ARG after 24 hours (28573±15947.07, p=0.0361) and 48 hours (45534.34±15916.60, p=0.0123) when compared to MSC cultivated alone. A very low, albeit significant 2-fold increase, was seen in Il4i1 expression after 48 hours (2.05±0.15 versus MSC alone, p=0.0102). Hdc expression was found to be significantly induced at 24 hours (18.30±7.50 versus MSC alone, p=0.0221), but declined at 48 hours (6.77±2.77). Ido mRNA levels were undetectable regardless of stimulation by Con A- activated splenocytes.

Therefore, MSC responded to an acute inflammatory microenvironment by upregulating EAA consuming enzymes, which have been implicated in immunological tolerance. The NOS2 response appeared to be the most prominent, followed by HDC and ARG2.

Figure 3.2. EAA consuming enzymes in MSC are differentially regulated by Con A-activated splenocytes. C57BL/6 MSC were cultured alone or co-cultured with naïve or Con A-activated splenocytes separated by a transwell membrane for 24 and 48 hours. After the time periods indicated, mRNA isolated from MSC were assayed by qRT-PCR. Data are shown as normalised to the Hprt1 housekeeping gene, and compared to the baseline expression in MSC alone. Each bar indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005 Unpaired t test. Arg1, arginase 1; Arg2, arginase 2; Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1; Il4i1, interleukin 4 induced 1; Nos1, nitric oxide synthase 1; Nos2, nitric oxide synthase 2.

3.2.3 Regulation of MSC immunosuppressive functions by inhibitors of EAA consuming enzymes

Having identified Nos2, Hdc, and Arg2 as the most abundant transcripts in MSC exposed to inflammation, the role of these molecules in affecting the immunomodulatory ability of MSC was investigated. Con A-stimulated splenocytes were cultivated in physical contact with different numbers of MSC in the presence or absence of pharmacological inhibitors of the three enzymes (1400W for NOS2, α- methyl-DL-histidine dihydrochloride for HDC and BEC for ARG1/ARG2) and cell proliferation was evaluated after 3 days. A titration assay was performed to evaluate the cell toxicity of inhibitors (Figure 3.3).

Remarkably, inhibition of NOS2 reduced MSC anti-proliferative activity although this was particularly evident at low MSC:splenocyte ratios (Figure 3.4). At 1:80, NOS2 inhibition only marginally impaired MSC anti-proliferation (80% of controls) (p=0.0043). The effect was much more prominent when the MSC:splenocyte ratio was 1:160 or 1:320, with the percentage of suppression compared to MSC in the absence of NOS2 inhibitor 40.40±10.40 (p=0.0044) and 10.46±0.89 (p<0.0001), respectively.

Con A-activated splenocytes were cultured with MSC for 3 days at an MSC:splenocyte ratio of 1:40, 1:80, 1:160, 1:320, as indicated. The ratios were chosen according to a range within which the anti-proliferative effect of MSC was demonstrated to be potent. The NOS2 inhibitor, 1400W, was tested at 2, 4, and 6μM. Inhibition of NOS2 remarkably reduced MSC anti-proliferative activity, although this was particularly evident at low MSC:splenocyte ratios (Figure 3.4). At a concentration of 2μM, a 20% decrease in inhibition was observed in the 1400W-treated, compared to the untreated group, at an MSC:splenocyte ratio of 1:320 (p=0.0051). When at 4μM, 1400W was more effective in blocking the anti-proliferative effect of MSC. At an MSC:splenocyte ratio of 1:80, a 60% reduction in MSC-mediated inhibition was obtained in the presence

81 of 1400W, whereas the proliferation was completely blocked by MSC (p=0.0046). The effect was much more prominent when the MSC:splenocyte ratio was 1:160 or 1:320 as the percentage of suppression compared to MSC in the absence of NOS2 inhibitor was 10.46±0.89 (p=0.0001) and 12.16±4.65 (p<0.0335), respectively. At 6μM, the decreased level of inhibition of splenocyte proliferation by 1400W, as compared to a full inhibition mediated by MSC, was 6-fold, 27-fold, and 58-fold at MSC:splenocyte ratios of 1:40, 1:160, 1:320, respectively (p=0.0032, p=0.0019, p=0.0023 respectively).

The effect of inhibiting HDC by α-methyl-DL-histidine dihydrochloride was then evaluated (Figure 3.5). At 1μM, impaired suppression of splenocytes was observed at an MSC:splenocyte ratio of 1:160 and 1:320 (95.68±4.79 versus 83.81±4.51, p=0.0350; and 69.12±1.79 versus 50.20±9.83, p=0.0302, untreated group versus treated group ratios of 1:160 and 1:320, respectively) but at a much lower degree than when NOS2 was inhibited. A similar pattern was observed 1.5μM of HDC inhibitor, with the suppression of splenocyte proliferation particularly evident at MSC:splenocyte ratios of 1:160 and 1:320 (p=0.0295, p=0.0066 respectively). However, increasing the dose of HDC inhibitor to 3μM did not produce any significant change. The negligible effect of HDC could be explained either by it having an intrinsically marginal role in immunosuppression and/or by the prominent effect of NOS2.

The anti-proliferative activities of ARG1 and ARG2 were then examined. As shown in Figure 3.6, the inhibition of ARG1 and ARG2 with BEC did not produce any effect on the anti-proliferative activity of MSC in the Con A-induced proliferation assay. When the dose of BEC was increased up to 250 μM, however, the proliferation of splenocytes in control cultures was impaired to 73.11±6.44, making the concentration not evaluable (data not shown). These data suggest that the anti-proliferative effect of MSC is independent of the catalytic activities of ARG1 and ARG2.

Figure 3.3. The dose of NOS2, HDC, and ARG inhibitors applied in the anti-proliferative assay is not cytotoxic. (A) 1400W was used as a NOS2 inhibitor at 2, 4, and 6µM; (B) α-methyl-DL-histidine dihydrochloride was used as an HDC inhibitor at 1, 1.5, and 3µM; (C) BEC was used as an ARG1/ARG2 inhibitor at 0.25, 1.25 and, 2.5µM. Splenocytes were plated at 5x105 cells/well and stimulated with 3µg/ml of Con A in the presence or absence (control) of inhibitor at the concentrations indicated for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. One representative experiment out of 3 independent experiments is shown. Proliferation was assessed by [3H] thymidine uptake, and the results are presented as percentage of inhibition, obtained using the formula:

{1-[(MSCSplConA)-(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD of cultures in triplicate.

Figure 3.4. NOS2 inhibition by 1400W blocks the anti-proliferative activity of MSC. Splenocytes were plated at 5x105 cells/well and stimulated with 3µg/ml of Con A in the presence or absence (controls) of C57BL/6 MSC at the ratio indicated. Cells were incubated with or without the NOS2 inhibitor, 1400W at various concentrations (2, 4, and 6μM) for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. One representative experiment out of 3 independent experiments is shown. Proliferation was determined by [3H] thymidine uptake, and the results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)-(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD of cultures in triplicates. *p < 0.05 ** p < 0.005 *** p < 0.001; Unpaired t test.

Figure 3.5. HDC inhibition by α-methyl-DL-histidine dihydrochloride counteracts the anti- proliferative activity of MSC. Splenocytes were plated at 5x105 cells/well and stimulated with 3µg/ml of Con A in the presence or absence (controls) of C57BL/6 MSC at the ratio indicated. Cells were incubated with or without the HDC inhibitor, α-methyl-DL-histidine dihydrochloride at various concentrations (1, 1.5, and 3μM) for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. One representative experiment out of 3 independent experiments is shown. Proliferation was determined by [3H] thymidine uptake, and results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)-(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD of triplicates of cultures. *p < 0.05; Unpaired t test.

Figure 3.6. ARG inhibition by BEC has no effect on the anti-proliferative activity of MSC. Splenocytes were plated at 5x105 cells/well and stimulated with 3µg/ml Con A in the presence or absence (controls) of C57BL/6 MSC at the ratio indicated. Cells were incubated with or without the ARG inhibitor, BEC, at various concentrations (0.25, 1.25, and 2.5μM) for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. One representative experiment out of 3 independent experiments is shown. Proliferation was determined by [3H] thymidine uptake, and the results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)-

(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD of triplicates of cultures.

3.2.4 Cytokine priming selectively skews EAA-consuming enzyme expression in MSC

Depending on specific TLR-agonist engagement, human MSC can be polarised towards a pro- or anti-inflammatory phenotype, exhibiting distinct abilities with respect to migration, invasion, and production of immunomodulatory factors [83, 88, 443]. To extend the polarisation paradigm observed to murine MSC, as well as to target the potential mechanisms by which the EAA consuming enzymes are induced by activated splenocytes, the expression of EAA consuming enzymes in response to a variety of cytokines involved in the inflammatory environment was investigated.

For this purpose, TNF-α and IFN-γ, either alone or in combination, were used as pro- inflammatory cytokines for the activation of MSC. IL-4 and IL-13 were used as anti- inflammatory cytokines. For TLR-priming, poly (I:C) and LPS were used as TLR3 and TLR4 agonists, respectively. After 24 or 48 hours of stimulation, changes in the MSC expression of candidate enzymes were evaluated. mRNA transcript levels of Arg1, Arg2, Nos1, Nos2, Il4i1, Hdc, and Ido were studied by using qRT-PCR. The concentrations of cytokines were chosen to induce the most significant effect on enzyme expression, according to the literature [88, 211, 444, 445]. In particular, concentrations of TNF-α and IFN-γ were chosen according to their abilities to induce Nos2 and Hdc, respectively (Figure 3.7). In addition to MSC generated from C57BL/6 (B6) mice, MSC generated from BALB/c mice were included in order to assess the potential similarities and differences in EAA consuming enzyme profiles between different strains.

Figure 3.7. Dose-dependent effects of TNF-α and IFN-γ on Nos2 and Hdc mRNA expression levels in MSC. C57BL/6 MSC were treated with different concentrations of cytokines: TNF-α (0.1, 1, 10, and 20 ng/ml), IFN-γ (5, 10, 20, and 40 ng/ml) for 24 hours. mRNA isolated from MSC was assayed by qRT- PCR. Data are shown as normalised to the housekeeping gene, Hprt1, and compared to baseline expression in untreated MSC. Each bar indicates the mean ± SD from three independent experiments. Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1; Nos2, nitric oxide synthase 2.

After 48 hours of exposure to the anti-inflammatory cytokines IL-4 and IL-13, B6 MSC showed a significant increase, albeit small, in the expression levels of Arg1 (1.55±0.15, p=0.0033 and 3.31±0.79, p=0.0074 versus untreated MSC, respectively) (Figure 3.8). Interestingly, Arg1 and Arg2 were modulated differently. The expression of Arg2 was highly up-regulated by pro-inflammatory cytokines and TLR4 agonists. At 24 hours, a significant increment was documented in Arg2 expression in MSC treated with TNF-α alone, or in combination with IFN-γ (11.33±2.67, p=0.0053 and 131.93±29.41, p=0.0413 versus untreated MSC, respectively). At 48 hours, a synergistic effect was observed upon treating MSC with TNF-α and IFN-γ to induce Arg2 expression, while treatment with LPS alone was also effective in up-regulating Arg2 (33.63±2.83, p=0.0002, and 3.34±1.24, p=0.0310 versus untreated MSC, respectively). An identical pattern, but at a much higher level, was observed for Nos2 transcript responses. MSC exposed to TNF-α upregulated Nos2 after both 24 and 48 hours (306.34±70.80,

88 p=0.0038 and 1267.15±878.02, p=0.0084 versus untreated MSC, respectively). Although IFN-γ did not induce any significant increase (64.66±48.26 and 71.33±53.87 versus untreated MSC at 24 hours and 48 hours, respectively), it synergised markedly with TNF-α (8194.11±2505.75, p=0.0039, and 16525.20±8963.19, p<0.0001 versus untreated MSC at 24 hours and 48 hours, respectively). Similarly to Arg1, the up- regulation of Il4i1 was mediated by IL-4 and IL-13 at 48 hours (1.32±0.10, p=0.0454, and 2.63±0.05, p=0.0006 versus untreated MSC, respectively). The expression level of Hdc was induced exclusively by IFN-γ after 24 hours (48.92±22.59, p=0.0213 versus untreated MSC). Nos1 was expressed at a relatively low level and could not be up- regulated in the presence of any cytokine. The expression level of Ido in untreated MSC was undetectable, whereas in the presence of IFN-γ, its expression was detectable at the qRT-PCR (~33 cycles) limit. Overall, the pro-inflammatory cytokines represent a more potent stimulus in inducing the expression of EAA consuming enzymes in MSC.

The EAA consuming enzyme expression profile of BALB/c MSC after cytokine treatments is shown in Figure 3.9. Arg1 was also up-regulated by anti-inflammatory cytokines, including IL-4 and IL-13 after 24 hours (9.39±6.43, and 8.57±1.08, p=0.0102 versus untreated MSC, respectively). This increase in Arg1 expression was also observed after 48 hours of IL-4 and IL-13 stimulation (20.25±21.07, 11.10±6.53 versus untreated MSC, respectively). In contrast, stimulation with TNF-α alone and in combination with IFN-γ up-regulated the expression of Arg2 after 24 and 48 hours, but such variation was not significant (25.09±19.00 and 18.43±10.99 versus untreated MSC at 24 hours, respectively; 94.44±129.85 and 126.97±143.65 versus untreated MSC at 48 hours, respectively). The mRNA levels were relatively low and no significant difference was observed after any cytokine. In contrast to Nos1, Nos2 was highly and significantly up-regulated when MSC were exposed to TNF-α alone and a similar level of expression was obtained upon TNF-α and IFN-γ co-treatment after 24 hours (2469.95±897.74, p=0.0089 and 2412.48±763.64, p=0.0054 versus untreated MSC, respectively). Furthermore, the effect of TNF-α on Nos2 expression was preserved after 48 hours, as the expression level was similar to that observed at 24 hours (2510.91±1347.74 versus untreated MSC). A significant increase in Nos2 was also documented following 48 hours of stimulation with LPS (529.30±291.96, p=0.0414 versus untreated MSC). The expression of Il4i1 was induced by TNF-α after 24 and 48

89 hours (8.07±7.88 and 6.84±8.00 versus untreated MSC), whereas its expression remained unchanged when MSC were treated with any of the other cytokines. In contrast to Arg2 and Nos2, the expression levels of Hdc were induced exclusively by IFN-γ with a peak after 24 hours (48.92±22.59, p=0.0213 versus untreated MSC), which declined at 48 hours. Despite stimulation by IFN-γ, untreated MSC expressed low levels of Ido, which were undetectable using qRT-PCR.

Due to their genetic differences, it has been shown that B6 and BALB/c mice express different immune responses in normal and pathological states [446]. Therefore, the expression of EAA consuming enzymes modulated by cytokine stimulations in MSC derived from these two strains were compared. The first noticeable difference was that IL-4 and IL-13 up-regulated Arg1 expression in BALB/c MSC, with a 10-fold increase, compared to the levels observed in B6 MSC (9.39±6.43 in BALB/c MSC against 0.78±002 in B6 MSC, 8.57±1.08 in BALB/c MSC against 0.80±0.27 in B6 MSC, after IL-4 and IL-13 stimulation, respectively). Another significant difference between the two strains was the synergistic effect documented in Arg2 and Nos2 expression after co-treatment with TNF-α and IFN-γ. After exposure to TNF-α and IFN-γ, a 130-fold increment in the expression of Arg2 was induced in B6 MSC in comparison with the 10-fold increase induced by TNF-α alone (p=0.0413 and p=0.0053, respectively). This effect was not observed in BALB/c MSC in which the levels of Nos2 induced by co- treatment with TNF-α and IFN-γ were similar to with TNF-α alone (18.43±11.00 and 25.01±19.00 versus untreated MSC, respectively). The synergistic effect mediated by TNF-α and IFN-γ co-treatment was also absent from Nos2 induction in BALB/c MSC. The levels of Nos2 expression in BALB/c MSC were similar when treated with TNF-α alone or in combination with IFN-γ (2469.95±897.74, p=0.0089 and 2412.48±763.64, p=0.0054 versus untreated MSC, respectively). The other EAA consuming enzymes showed a similar regulation pattern in both strains after cytokine treatments. As differences were observed between the two strains, only MSC generated from B6 mice were employed in the rest of the study, avoiding potential variation when compared to Nos2-/- MSC and Tnfr1/r2-/- MSC, which were both generated from mice in a B6 background.

Figure 3.8. Profile of EAA consuming enzyme expression in MSC isolated from C57BL/6 mice. C57BL/6 (B6) MSC were treated with the following cytokines: IL-4 (20 ng/ml), IL-13 (20 ng/ml), IFN- γ (20 ng/ml), TNF-α (20 ng/ml), poly (I:C) (1 μg/ml) and LPS (10 ng/ml) either alone or in combination (IFN-γ and TNF-α) for 24 and 48 hours. After the time periods indicated, mRNA isolated from MSC were assayed by qRT-PCR. Data shown are normalised to the housekeeping gene, Hprt1, and compared to baseline expression in untreated MSC. Each bar indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005 ***p < 0.001 ****p < 0.0001 Unpaired t test. Arg1, arginase 1; Arg2, arginase 2; Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1; Ido, indoleamine 2,3-dioxygenase; Il4i1, interleukin 4 induced 1; Nos1, nitric oxide synthase 1; Nos2, nitric oxide synthase 2. 91

Figure 3.9. Profile of EAA consuming enzyme expression differs in MSC generated from BALB/c mice. BALB/c MSC were treated with the following cytokines: IL-4 (20 ng/ml), IL-13 (20 ng/ml), IFN- γ (20 ng/ml), TNF-α (20 ng/ml), poly (I:C) (1 μg/ml) and LPS (10 ng/ml) either alone or in combination (IFN-γ and TNF-α) for 24 and 48 hours. After the indicated time periods, mRNA isolated from MSC were assayed by qRT-PCR. Data shown are normalised to the housekeeping gene, Hprt1, and compared to baseline expression in untreated MSC. Each bar indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005 Unpaired t test. Arg1, arginase 1; Arg2, arginase 2; Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1; Ido, indoleamine 2,3- dioxygenase; Il4i1, interleukin 4 induced 1; Nos1, nitric oxide synthase 1; Nos2, nitric oxide synthase 2

3.2.5 NOS2 deficiency leads to the impairment of the immunosuppressive functions of MSC

In order to confirm the data observed with the NOS2 inhibitor, MSC generated from Nos2-/- mice were tested for their anti-proliferative functions. Firstly, changes in the expression of NOS2 protein (130 kDa) between WT MSC and Nos2-/- MSC were evaluated by Western blot analysis (Figure 3.10). Consistent with what had been observed in the modification of gene expression, NOS2 production was low in the untreated WT MSC, whereas the amount of NOS2 increased substantially in response to TNF-α and IFN-γ co-stimulation. NOS2 production in Nos2-/- MSC was abolished and could not be induced by cytokine stimulation. The protein expression of NOS2 in WT MSC and Nos2-/- MSC was further confirmed by confocal staining. WT MSC and Nos2-/- MSC stimulated with TNF-α and IFN-γ for 24 hours were stained by antibodies against NOS2, nuclear DNA content (DAPI) and tubulin (for cytoskeleton staining) (Figure 3.11). Confocal images revealed a marked distribution of NOS2 expression induced by co-stimulation with TNF-α and IFN-γ. Instead, such induction was not observed in Nos2-/- MSC, demonstrating their NOS2 deficiency.

Figure 3.10. The absence of NOS2 after cytokine stimulation confirmed the deficiency of NOS2 in Nos2-/- MSC. MSC were stimulated with TNF-α (20ng/ml) and IFN-γ (20ng/ml) for 24 hours. The cells were then harvested and the production of NOS2 was evaluated by western blotting. β-Actin was used as a loading control for all immunoblots.

Figure 3.11. Confocal analysis of NOS2 expression in MSC after stimulation by inflammatory cytokines. (A) WT MSC; (B) Nos2-/- MSC untreated (left panel), or treated with TNF-α (20ng/ml) and IFN-γ (20ng/ml) (right panel) for 24 hours were stained with DAPI (nucleus, blue), anti-α-tubulin (microtubules, green), and anti-NOS2 (nitric oxide synthase 2, red). Scale bar, 10 µm.

Further evidence supporting the impairment of MSC with the NOS2 inhibitor was that Nos2-/- MSC exhibited a much reduced ability to inhibit Con A-induced cell proliferation compared to WT MSC (Figure 3.12A). At all MSC:splenocyte ratios tested, the inhibition decreased to less than 20% (p<0.001). A significant difference in the anti-proliferative effect of Nos2-/- MSC and WT MSC was obtained at a ratio of 1:640 when the dose of WT MSC was not effective at inhibiting splenocyte proliferation (data not shown).

Since the expression of Hdc was found to respond exclusively to IFN-γ stimulation and it emerged as an alternative potential pathway in functional assays, the question of whether the anti-proliferative ability could be restored in Nos2-/- MSC by up-regulating their HDC expression with IFN-γ pre-treatment was investigated (Figure 3.12B). MSC were pre-treated with IFN-γ for 24 hours, and the anti-proliferative activity revealed a minor increase in the inhibition in IFN-γ pre-treated Nos2-/- MSC compared to untreated Nos2-/- MSC at low MSC:splenocyte ratios (25.15±3.87 versus 16.32±3.42, IFN-γ pre- treated versus untreated Nos2-/- MSC at 1:320). Nevertheless, no difference was obtained in the anti-proliferative ability of MSC from pre-treatment with IFN-γ. Therefore, I evaluated the level of Hdc expression in the untreated or IFN-γ pre-treated Nos2-/- MSC after exposure to naïve or Con A-activated splenocytes for 24 hours (Figure 3.12C). After exposure to activated splenocytes, Hdc expression was induced to a higher degree in IFN-γ treated Nos2-/- MSC compared to untreated Nos2-/- MSC (38.18±11.78 versus 34.54±10.64), however, this difference was not statistically significant. Notably, even without activation by IFN-γ, Hdc expression was significantly induced in Nos2-/- MSC at a greater level than WT MSC following exposure to Con A-activated splenocytes (34.54±10.64 versus 5.88±1.87, Nos2-/- MSC versus WT MSC, p=0.0078). These results suggest that the residual immunosuppressive activity of Nos2-/- MSC, might be mediated by HDC.

Figure 3.12. Deficiency in NOS2 eliminates the anti-proliferative activity of MSC, whereas the presence of HDC induced by IFN-γ partially reverses the effect. (A) WT or Nos2-/- MSC; (B) Untreated or IFNγ pre-treated (20ng/ml) Nos2-/- MSC were plated at 1.25x104 cells/well, with 1:2 serial dilutions, and splenocytes were added at 5x105 cells/well and stimulated with 3μg/ml of Con A for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake, and results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)- (MSCSpl)/(SplConA-Spl)]}*100. (A) Each column represents the mean ± SD from three independent experiments (B) Each column represents the mean ± SD of cultures in triplicate, a representative experiment out of two independent experiments is shown. (C) HDC expression in untreated or IFN-γ pre-treated Nos2-/- MSC after stimulation. WT and Nos2-/- MSC were pre-treated with IFN-γ (20ng/ml) for 24 hours. The cells were washed intensively before co-culture with naïve or Con A-activated splenocytes separated by a transwell membrane for 24 hours. mRNA isolated from MSC was assayed by qRT-PCR. Data are normalised to the housekeeping gene, Hprt1, and compared to baseline expression in WT MSC. Each bar indicates the mean ± SD from three independent experiments. *p < 0.05 ** p < 0.005 *** p < 0.001; Unpaired t test. Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1

NOS2 has been demonstrated to be the most potent among the three known isoforms of NOS due to its ability to produce sustained NO concentrations in a μM range [447]. Notably, NO is known to be produced by macrophages as a potent T cell suppressant [241, 448]. Since up-regulation of NOS2 has been reported to evoke the formation of NO, the ability of Nos2 deficient MSC to produce NO was investigated. It has been observed that activated T cells can secrete small amounts of NO and so to exclude this possibility, the culture medium of splenocytes was performed as follows: culture medium collected from Con A-activated splenocytes was used to stimulate MSC for 24, 48, 72 hours, and the supernatants were assayed for the presence of nitrate and nitrite as an estimate of NO production (Figure 3.13).

The supernatants from splenocytes cultured in the absence or presence of Con A stimulation had a low NO concentration: 1.73±1.60 μM and 1.96±1.96 μM, respectively (per 1.25x105 cells). Consistent with the production of NOS2, the addition of activated splenocyte supernatant triggered MSC-mediated NO production significantly after 24 hours (15.67±2.48 (μM/1.25x106 cells), p=0.0017 versus supernatants from Con A-induced splenocytes). The up-regulation of NO production by MSC occurred in a time-dependent manner after exposure to activated splenocyte supernatant (24.73±9.21 and 28.00±10.15 (μM/1.25x105 cells), after 48 and 72 hours, respectively). The level of NO remained unchanged when MSC were exposed to non- activated splenocyte supernatant (data not shown). When Nos2−/− MSC were stimulated by activated splenocyte supernatant, a limited increase in their NO production was observed at all the time points investigated (2.60±2.00, 2.43±1.91, and 2.83±1.81 (μM/1.25x105 cells) at 24 hours, 48 hours, and 72 hours, respectively). These results suggest that NOS2 induced by MSC was the main source of NO after exposure to activated splenocytes.

Figure 3.13. Abolished NO production is observed in Nos2-/- MSC compared to WT MSC after stimulation. MSC were stimulated with supernatants of cultures from concanavalin A (Con A)-activated murine splenocytes for 24, 48 and 72 hours. Griess reagent was used to measure total nitrate/nitrite concentration in the medium as an estimate of NO production. Each column indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005; Unpaired t test.

3.2.6 Eliminating the effect of TNF-α partially impaired the immunosuppressive activities of MSC

Considering the key role of TNF-α stimulation in driving Nos2 expression, whether TNF-α was also the main molecule accounting for the anti-proliferative promoting ability of ConA-stimulated splenocytes was also investigated. Therefore, we interrogated MSC generated from TNF receptor 1/receptor 2 deficient (Tnfr1/r2-/-) mice for their functional activity.

Firstly, as the expression of Arg2 and Nos2 has been shown to be promoted by TNF-α, the expression profile of these two molecules was assessed in TNF receptor 1/receptor 2 deficient (Tnfr1/r2-/-) MSC by qRT-PCR. As observed in Figure 3.14, the synergistic effect of TNF-α and IFN-γ co-treatment in inducing Arg2 level was absent in Tnfr1/r2- /- MSC (23.23±6.6 versus 1.08±0.36, p=0.0005, and 29.48±10.40 versus 0.48±0.38, p=0.0014, WT MSC versus Tnfr1/r2-/- MSC at 24 and 48 hours, respectively). In response to Con A-activated splenocytes, the level of Arg2 was diminished by one half in Tnfr1/r2-/- MSC compared to WT MSC (1.99±1.95 versus 1.02±0.39, and 7.50±3.94 versus 4.36±4.89, WT MSC versus Tnfr1/r2-/- MSC at 24 and 48 hours, respectively). As expected, the Nos2 up-regulation induced by TNF-α was not observed in Tnfr1/r2- /- MSC. There was a significant decrease in Nos2 expression when Tnfr1/r2-/- MSC were co-cultured with Con A-stimulated splenocytes (2138.89±428.53 versus 396.26±65.38, p=0.0296, WT MSC versus Tnfr1/r2-/- MSC at 24 hours). However, the difference in Nos2 expression induced by activated splenocytes was not statistically significant between WT MSC and Tnfr1/r2-/- MSC due to variation. Surprisingly, the level of IFN- γ induced Hdc up-regulation was also significantly decreased in Nos2-/- MSC; from 30.33±8.08 to 6.88±4.95 at 24 hours (p=0.0049), and from 19.16±10.61 to 1.61±0.75 at 48 hours (p=0.0104).

Figure 3.14. Tnfr1/r2-/- MSC express different patterns of EAA consuming enzymes compared to WT MSC in response to inflammatory cytokines and Con A-activated splenocytes. MSC were treated with the following cytokines: TNF-α (20ng/ml), IFN-γ (20ng/ml) either alone or in combination, or co-cultured with Con A-activated splenocytes in the transwell for 24 hours and 48 hours. After the indicated time periods, mRNA isolated from MSC was assayed by qRT-PCR. Data are normalised to the housekeeping gene, Hprt1, and compared to baseline expression in WT MSC. Each bar indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005 ***p < 0.001; Unpaired t test. Arg2, arginase2; Hdc, histidine decarboxynase; Hprt1, hypoxanthine guanine phosphoribosyl transferase 1; Nos2, nitric oxide synthase 2

100

The level of NOS2 in Tnfr1/r2-/- MSC was assayed by Western blot analysis following stimulation by activated splenocytes. As shown in Figure 3.15, the production of NOS2 protein was markedly reduced in Tnfr1/r2-/- MSC. As additional evidence, the amount of NO produced by Tnfr1/r2-/- MSC (Figure 3.16) was measured. Culture medium collected from Con A-activated splenocytes was used to stimulate MSC for 24, 48 and 72 hours, and, supernatants were assayed for the presence of nitrate and nitrite as an estimate of NO production. A significant decrease in NO secretion was observed in Tnfr1/r2-/- MSC in comparison to WT MSC at 24 hours (7.55±2.19 versus 15.67±2.48 (μM/1.25x106 cells), Tnfr1/r2-/- MSC versus WT MSC, p=0.0021). At 48 hours, the decrease could still be identified without being statistically significant (11.45±6.29 versus 24.73±9.21, Tnfr1/r2-/- MSC versus WT MSC). Nonetheless, the amount of up- regulated NO in Tnfr1/r2-/- MSC was demonstrated to be significant compared to the control at 72 hours (14.90±3.11 versus 2.33±1.81 (μM/1.25x106 cells), Tnfr1/r2-/- MSC versus culture medium from activated splenocytes, p=0.0097).

Figure 3.15. Inflammation-induced NOS2 upregulation is impaired in Tnfr1/r2-/- MSC. MSC were co-cultured with Con A-activated splenocytes in the transwell for 24 hours. The cells were then harvested and the production of NOS2 was evaluated by Western blotting. β-Actin was a loading control in all immunoblots.

101

Figure 3.16. Inflammation-induced NO production is impaired in Tnfr1/r2-/- MSC. MSC were stimulated with supernatants from cultures of Con A-activated murine splenocytes for 24, 48 and 72 hours. Griess reagent was used to measure total medium nitrate/nitrite as an estimate of NO production. Each column indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005, NS, not significant. Unpaired t test.

102

The anti-proliferative ability of Tnfr1/r2-/- MSC was found to be impaired, but not to the same extent as that observed in Nos2-/- MSC (Figure 3.17). Compared to WT MSC, a significant reduction in splenocyte inhibition was documented at MSC:Splenocyte ratios from 1:40 to 1:160 with Tnfr1/r2-/- MSC (99.50±0.86% versus 78.32±5.09%, p=0.0021, and 97.92±3.58% versus 60.57±17.23%, p=0.0213, and 95.48±5.71 versus

43.34±11.27%, p=0.0020, WT MSC versus Tnfr1/r2-/- MSC at MSC:Splenocyte ratios of 1:40, 1:80, 1:160, respectively).

In addition to Tnfr1/r2-/- MSC, the role of TNF-α was tested using etanercept, a TNFα receptor antagonist, to block the effect of TNF-α in MSC-mediated anti-proliferation.

The addition of etanercept produced results consistent with those in Tnfr1/r2-/- MSC and a similar pattern of down-regulated inhibition was observed. In the presence of etanercept, 14.63%, 19.68%, 29.08%, and 24.49% less inhibition of activated splenocytes was observed with 1:40, 1:80, 1:160, and 1:320 MSC:Splenocyte ratios, respectively. Collectively, the results indicate that TNF-α is a key regulator of NOS2, but also suggest the existence of alternative concomitant pathways for the induction of

NOS2.

103

Figure 3.17. TNF-α plays a crucial role in the anti-proliferative activity of MSC. (A) WT or Tnfr1/r2- /- MSC (B) Untreated or treated WT MSC with 1μg/ml etanercept were plated at 1.25x104 cells/well, with 1:2 serial dilutions, and splenocytes were added at 5x105 cells/well and stimulated with 3μg/ml of Con A for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake, and results are expressed as percentages of inhibition, which was determined using the formula: {1-[(MSCSplConA)-

(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD from two independent experiments. *p < 0.05 **p < 0.005 ***p < 0.001 ****p < 0.0001; Unpaired t test.

104

3.2.7 The role of NF-κB as a key upstream mediator to induce NOS2 expression in MSC

NF-κB is a transcription factor that governs the expression of genes encoding cytokines, chemokines, and growth factors in response to infection and stress. There are five subunits within the NF-κB family: RelA (p65), RelB, c-Rel, and the precursor proteins, NF-κB1 (p105) and NF-κB2 (p100). In the inactivated form, NF-κB is found in the cytoplasm in a complex with an NF-κB inhibitor, IκB. Upon activation, IκB is phosphorylated and targeted by proteasomal degradation. In the canonical pathway, the phosphorylation of IκB is mediated by IKKβ, which allows NF-κB1 to translocate to the nucleus and then be processed into p50. In the non-canonical pathway, activation of IKKα by the NF-κB-inducing kinase (NIK) promotes the phosphorylation of IκB and results subsequently in the processing of NF-κB2 into p52 [449-451]. A variety of stimuli have been identified that induce the degradation of IκB and the concomitant activation of NF-κB through cytokine receptors and pattern-recognition receptors. Among these stimuli, TNF-α triggers the activation of NF-κB through both the canonical and non-canonical pathways and induces downstream gene transcription. NOS2 is one of the target genes regulated by NF-κB and in macrophages it has been demonstrated that LPS induced expression of NOS2 requires the activation of NF-κB. Considering the fact that the activation of NFκB can be regulated by TNF-α, while NF- κB serves as an upstream regulator of NOS2, whether enhanced NOS2 expression in MSC depends on the activation of NF-κB was investigated here.

105

To elucidate the potent molecular mechanism by which NOS2 expression was induced when MSC exert anti-proliferative activity, the role of NF-κB was evaluated. Firstly, the nuclear translocation of p65 was examined as an indication of NF-κB activity in MSC after TNF-α stimulation. As shown in Figure 3.18, the expression of p65 in the nucleus became abundant after this treatment, suggesting that the activation of NF-κB was induced by TNF-α. On the other hand, treatment with 10 μM of inhibitor Bay11- 7085, which suppresses IκB phosphorylation, was sufficient to block the activation of NF-κB since the translocation of p65 was not observed.

To demonstrate the linkage between NF-κB and NOS2 in MSC, direct evidence for the regulation of NOS2 expression by NF-κB activation and inhibition was sought. qRT- PCR revealed that by applying the IKKβ inhibitors, TPCA-1 and PF-026, or the IκB phosphorylation inhibitor, Bay11-7085, the induction of Nos2 expression was significantly diminished during co-culture with activated splenocytes. As observed previously, the expression of Nos2 was induced after MSC were exposed to Con A- activated splenocytes (2343.88±302.80, p=0.0082 versus untreated MSC) (Figure 3.19). Following NF-κB inhibition mediated by TPCA-1, BAY11-7085, and PF-026, Nos2 expression was reduced to a relatively low level and was not affected by activated splenocytes (4.79±1.67, 8.01±5.69, and 1.40±0.09 versus untreated MSC, respectively). Nonetheless, in the presence of ML-120B, the reduction in Nos2 expression did not occur during co-culture with activated splenocytes (1074.99±24.34, p=0.0003 versus untreated MSC).

In order to confirm the effect of NF-κB on NOS2 activity, the production of NO in MSC was measured in the presence of NF-κB inhibitors (Figure 3.20). The amount of NO produced by MSC increased after 24 hours of exposure to the supernatant collected from Con A-activated splenocytes (8.91±0.19 versus 14.39±0.97 (μM/1.25x106 cells), p=0.0162). In contrast, the production of NO in MSC exposed to activated splenocyte culture medium was found to remain at the same basal levels as for naïve MSC in the presence of NK-κB inhibitors (7.74±0.03 versus 8.29±0.03 (μM/1.25x106 cells), 7.97±0.16 versus 8.34±0.29 (μM/1.25x106 cells), naïve MSC versus MSC stimulated by supernatant of Con A-activated splenocytes in the presence of BAY11-7085 and PF- 026, respectively). Consistent with results obtained in qRT-PCR, MSC treated with ML-120B still produced significant amounts of NO in response to stimulation 106

(8.22±0.13 versus 12.78±0.45 (μM/1.25x106 cells), naïve MSC versus MSC stimulated by the supernatant of Con A-activated splenocytes in the presence of ML-120B, p=0.0054).

Figure 3.18. Effect of NF-κB inhibitor, BAY11-7085, on TNF-α induced nuclear translocation of p65 in MSC. MSC were pre-treated with BAY11-7085 (10 μM) for one hour. Cells were washed with PBS three times, then stimulated with 20 nM TNF-α for 24 hours. p65 expression in cytosolic and nuclear extracts of MSC was evaluated by Western blotting. β-Actin was used as the loading control for all immunoblots.

Figure 3.19. NF-κB signalling is essential for NOS2 expression in response to Con A-activated splenocytes. MSC were pre-treated with NF-κB inhibitors: TPCA-1 (10 µM), BAY-117085 (10 µM), ML-120B (10 µM) and Pfizer (10 µM) for one hour. Cells were washed with PBS three times before co- culture with Con A-activated splenocytes in the transwell for 24 hours. mRNA expression in MSC was detected using qRT-PCR. Data are shown as normalised to the housekeeping gene, hprt1, and compared to baseline expression in the WT MSC. Each bar indicates the mean ± SD from two independent experiments. **p < 0.005 ***p < 0.001; Unpaired t test. Nos 2, nitric oxide synthase 2. 107

Figure 3.20. NO production is suppressed by NF-κB inhibition. MSC were treated with NF-κB inhibitors: TPCA-1 (10µM), BAY-117085 (10µM), ML-120B (10µM) and PF-026 (10µM). They were then stimulated with supernatant from Con A-activated murine splenocytes cultures for 24 hours. Griess reagents were used to measure total nitrate/nitrite concentrations in the media as an estimate of NO production. Each column indicates the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005; Unpaired t test. NS, not significant.

According to the previous data obtained, the generation of NO is crucial to MSC- mediated anti-proliferation. Since the production of NO was completely abolished by the NF-κB inhibitors, BAY-117085 and PF-026, the effect of these inhibitors on MSC- mediated T cell proliferation was investigated. The results presented in Figure 3.21A illustrate that the anti-proliferative effect of MSC was eliminated when MSC were treated with either 10μΜ or 30μΜ BAY-117085. From MSC:Splenocyte ratios of 1:40 up to 1:320, the impairment in the proliferation of splenocytes was observed in a dose- independent manner (1.35±0.20%, 1.31±0.01%, 1.49±0.49%, and 9.29±5.13% at MSC:Splenocyte ratios of 1:40, 1:80, 1:160, and 1:320, respectively). Although PF-026 was shown to reduce the secretion of NO by MSC, it had a limited effect on diminishing the inhibition of MSC. As shown in Figure 3.21B, the inhibition of splenocytes was significantly reduced when co-cultured with 10μΜ PF-026-treated MSC at MSC:Splenocyte ratios of 1:40 (52.40±3.09% versus 64.35±8.60%, p=0.0155, PF-026- treated MSC versus WT MSC). Nonetheless, the anti-proliferative ability of 10μΜ PF- 026-treated MSC was as effective as WT MSC at MSC:Splenocyte ratios of 1:80 to

108

1:320. It was surprising that by treating MSC with 30μΜ of PF-026, the anti- proliferative effect of MSC was significantly enhanced at a MSC:Splenocyte ratio of 1:320 (79.76±4.72% versus 34.80±8.73%, p=0.0014, 30μΜ PF-026-treated MSC versus WT MSC).

Figure 3.21. NF-κB inhibition impairs the anti-proliferative ability of MSC. WT MSC were plated at 1.25x104 cells/well in the presence of inhibitor (A) BAY11-7085 or (B) PF-026 (10 and 30 μM). Splenocytes were added at 5x105 cells/well and stimulated with 3μg/ml of Con A for 3 days; 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake, and results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)-(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD from three independent experiments. *p < 0.05 **p < 0.005 ***p < 0.001 ****p < 0.0001; Unpaired t test.

109

3.2.8 The effect of MyD88 in regulating NOS2 expression in MSC

In response to various PAMPs derived from bacteria, viruses, fungi, and protozoa, TLRs are activated and subsequently trigger intracellular signalling pathways involved in adaptive immune response[52]. Downstream signalling through TLRs can be classified into two different pathways depending on the type of intracellular adaptor protein recruited, MyD88-dependent pathway and Toll-IL-1 receptor domain- containing adaptor-inducing IFN-β (Trif)-dependent pathway. The MyD88-dependent pathway is shared by all TLRs except TLR3, whereas the Trif-dependent pathway is only activated by TLR3 and TLR4 [58, 452, 453]. Both pathways can induce the activation of the mitogen-activated protein kinases (MAPKs) and NF-κB. It has been shown that stimulation by IL-1β and all other TLR ligands, except for TLR3 and TLR4 ligands, fails to promote MAPKs or NF-κB activation in Myd88-/- mice [454]. However, the activation of MAPKs and NF-κB is observed in Myd88-/- mice with delayed kinetics [455]. Thus, Trif represents an alternative pathway, which contributes to MAPKs and NF-κB induction. Interestingly, TLR4-induced MyD88 and NOS2 up- regulation have been demonstrated to protect cardiomyocytes against apoptosis [456]. In addition, MyD88 has been shown to induce NOS2-mediated NO production in macrophages after their activation by IL-1β. Such a mechanism has been observed to be involved in host resistance against infection, when the higher production of NO restricts the replication of Leishmania [457]. Based on these notions, the possibility of MyD88 being one of the up-stream mediators involved in MSC-induced NOS2 up- regulation and anti-proliferative activity was investigated here.

110

Genotypic analysis of Myd88-/- MSC was performed by PCR to confirm the gene deficiency. As shown in Figure 3.22, MyD88 DNA was absent in Myd88-/- MSC, whereas DNA specific for the knockout alleles was undetectable in WT MSC. Since previous studies have indicated that signalling through either the TLR4 or IL-1 receptor was sufficient to activate MyD88 and subsequently to trigger NOS2 expression [457], NOS2 gene expression in Myd88-/- MSC was evaluated, revealing that it was significantly reduced in comparison to WT MSC after treatment of LPS and IL-1β (Figure 3.23). Exposure to LPS promoted a high amount of Nos2 expression in WT MSC, whereas a deficiency in MyD88 resulted in a significant reduction in Nos2 expression (1188.01±312.12 versus 1.78±0.73, WT MSC versus Myd88-/- MSC, p=0.0329). The up-regulation in Nos2 expression by IL-1β was 2000 fold versus untreated WT MSC, whereas this effect was not observed in Myd88-/- MSC (0.81±0.13).

An anti-proliferative assay was applied in order to evaluate the functional defect that was due to NOS2 impairment in Myd88-/- MSC. At an MSC:Splenocyte ratio of 1:80, a noticeable decrease was observed in Myd88-/- MSC-mediated inhibition (99.97±0.72 versus 72.05±10.15, WT MSC versus Myd88-/- MSC). The difference became more prominent at lower MSC:Splenocyte ratios (98.58±0.03 versus 60.02±9.93, p=0.0316 and 86.05±4.70 versus 38.32±10.97, p=0.0299, WT MSC versus Myd88-/- MSC at ratios of 1:160 and 1:320, respectively). Therefore, MyD88 appeared to be one of the up-stream effectors for NOS2 up-regulation in MSC exposed to inflammation, and the lack of MyD88 mediated signalling partially impaired the anti-proliferative effect of MSC.

111

Figure 3.22. Genotypic analysis of Myd88-/- MSC. The expected sizes of the PCR products of wild type, knockout allele genotypes for MyD88 are shown in the panel on the left. The top lane indicates the resource from which the DNA sample was generated. The bottom lanes indicate the primers that were used for either WT or MyD88-

Figure 3.23. NOS2 expression is impaired in Myd88-/- MSC in response to inflammatory cytokines and TLR4 ligand. MSC were treated with IL-1β (20ng/ml) or LPS (10ng/ml) for 24 hours. mRNA expression in MSC was assayed by qRT-PCR. Data shown are normalised to the housekeeping gene, hprt1, and compared to baseline expression in the WT MSC. Each bar indicates the mean ± SD from two independent experiments. *p < 0.05; Unpaired t test; Nos2, nitric oxide synthase 2.

112

Figure 3.24. Myd88-/- MSC exhibit reduced anti-proliferative activity. WT or Myd88-/- MSC were plated at 1.25x104 cells/well, with 1:2 serial dilutions, and splenocytes were added at 5x105 cells/well and stimulated with 3μg/ml of Con A for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake, and the results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCSplConA)-

(MSCSpl)/(SplConA-Spl)]}*100. Each column represents the mean ± SD from two independent experiments. *p < 0.05; Unpaired t test.

113

3.3 Discussion

In the first part of my project, I generated MSC that exhibited an immunophenotype consistent with the current criteria. They were negative for leukocyte marker CD45 and expressed PDGFR, Sca-1, and CD90 (Figure 3.1). The co-expression of PDGFR and Sca-1 has been described as a marker of MSC, which identifies the population in which the progenitor activity resides [458]. However, these markers do not identify any specific property or characterise the differentiation stage along the mesenchymal lineage. Furthermore, despite indications in favour of their progenitor status, there is no convincing evidence that this property is associated with immunomodulatory activity. Therefore, my studies provide a detailed characterisation of these MSC based on their immunoregulatory functions.

The immunosuppressive effects of MSC have been correlated to the secretion of soluble factors. Among the potential candidates examined, IDO [181], NOS2 [79, 167], heme oxygenase (HO)-1 [459], TGF-β [460], PGE2 [101], IL-6 [461], and human leukocyte antigen (HLA)-G5[126] are all considered to be involved in MSC-mediated immune regulation. Amongst the various mechanisms, EAA metabolism may serve as the principal mechanism exploited by MSC to alter the microenvironment and inhibit the proliferation of target cells. This is consistent with the notion that MSC immunosuppression only acts locally. I evaluated the profile of EAA-consuming enzymes in MSC exposed to splenocytes (Figure 3.2). MSC up-regulated a set of EAA only when in co-culture with activated, but not with resting, splenocytes. In accordance with previous studies showing that T cell activation is required by MSC for their acquired immunosuppressive function [79], this result highlights the role of the microenvironment in modulating MSC activity.

I observed that MSC up-regulated both ARG1 and ARG2 isoforms in response to activated-splenocytes, but Arg2 transcript levels were increased in higher amounts (Figure 3.2). The activation of ARG1 has been demonstrated to induce T-cell anergy and facilitate tumour growth in MDSC and DC [205, 210, 462]. Similar activity has been reported in rodent macrophages, which represent a major population with respect to governing arginase activities during the immune response [463]. Few studies have

114 examined the role of ARG2. One study showed that, in response to LPS stimulation, ARG2 promoted pro-inflammatory cytokine production in macrophages via mitochondrial ROS [464]. Furthermore, the accumulation of ARG2-expressing macrophages was associated with atherosclerotic lesions. Interestingly, ARG2 is mediated by murine renal carcinoma cell lines to induce arginine deprivation and down- regulate CD3ζ expression in co-cultured T cells [207]. My results indicate for the first time that the increment in Arg2 in MSC can be regulated by pro-inflammatory cytokines.

Apart from ARG, NOS constitutes another pathway that mediates arginine metabolism and suppresses immune cell proliferation. I observed a massive increment in Nos2 expression in MSC cultivated with activated splenocytes, whereas the level of Nos1 remained unchanged (Figure 3.2). Among three isoforms of NOS, NOS2 is constitutively active once induced and its activation plays a fundamental role in immune regulation. NOS expression is a common feature of genuine immune cells including dendritic cells, NK cells, mast cells, monocytes, macrophages and neutrophils [465, 466]. In addition, NOS can be mediated by niche-composing cells to obtain a diversity of functions. It has been observed that mesangial cells regulate NOS2 activity in a similar way to macrophages [467]. In addition to macrophages and other inflammatory cells, NOS2 activity in fibroblasts contributes to regulating inflammation and wound healing [468]. NOS3 activity in osteoblasts involves the regulation of cell proliferation and differentiation, suggesting its relevance to bone maintenance [469]. In endothelial cells, the expression of NOS3 is fundamental for maintaining cardiovascular homeostasis and ameliorating vascular disease [470]. NO production by vascular smooth muscle cells was surmised to prevent atherosclerosis and restenosis by inhibiting smooth muscle cell proliferation and contraction, as well as by inhibiting leukocyte and platelet adhesion [471]. NOS2 activity in hepatocytes is considered to protect them from TNF-α mediated apoptosis [472].

HDC represents an exclusive source generating histamine production, which plays an important role in immunomodulation [473]. The biological importance of HDC and histamine is surmised since HDC is widely expressed by a variety of cells such as haematopoietic progenitors, macrophages, platelets, dendritic cells and T cells [474]. In MSC, histamine was first suggested to regulate the osteogenic differentiation 115 pathway [475]. Later, Nemeth and colleagues established the effect of histamine in MSC by showing that inducible IL-6 production is associated with the binding of histamine to H1 and H2 receptors on MSC [476]. Apart from this, little is known about the regulation of HDC activity and the production of histamine by MSC. For the first time, my findings have demonstrated that the expression of Hdc can be highly up- regulated by MSC following the stimulation of activated splenocytes (Figure 3.2).

Through depleting local tryptophan and inducing tryptophan metabolite accumulation, IDO has been assigned a central role in suppressing the T-cell response at sites of inflammation and in the microenvironment of tumour [293, 298, 306]. Dendritic cells [298], macrophages [292], tumour cells [309], and specialized epithelial cells (e.g., syncytiotrophoblasts in the placenta) [477] have been documented to modulate IDO activity. As observed in human MSC, IDO-mediated tryptophan catabolism leads to T- cell inhibition in MLRs [166]. Here, IDO was undetectable in the murine MSC/Splenocytes co-culture system (Figure 3.2) despite the large body of evidence suggesting that IDO plays a crucial role in tolerogenic dendritic cells in mice. Even though human IDO shares 62% sequence homology with murine IDO, it seems that the regulation of IDO production by inflammatory stimulants is variable between species [478, 479]. To clarify this, a recent study employed either NOS2 or IDO inhibitors to investigate immunoregulatory relevance in MSC of different species. Upon stimulation from activated splenocytes or peripheral blood mononuclear cells, the immunosuppressive effect of MSC from mouse, rat, hamster, and rabbit was dependent on NOS2. In contrast, IDO was utilised by MSC to perform immunosuppression in monkey, pig, and in human [480]. This observation may account for the lack of IDO activation in murine MSC. In fact, human MSC assessed in our laboratory clearly respond to IFN-γ with vigorous IDO up-regulation (Galleu et al, in preparation).

My results demonstrated a small increase in the level of Il4i1 in MSC after exposure to activated splenocytes for 48 hours (Figure 3.2). Phenylalanine oxidase IL4i1 is primarily expressed by activated macrophages and dendritic cells after stimulation by pro-inflammatory mediators [344, 481]. In addition, Il4i1 has also been detected in B lymphocytes, albeit at much lower levels [348].

116

Since abundant production of Nos2, Hdc, and Arg2 was observed in MSC after stimulation, their functional importance was evaluated by quantifying the ability of MSC to inhibit lymphocyte proliferation in the presence of enzyme antagonists. Consistent with previous findings [79, 167], the inhibitory ability of MSC was reduced by approximately 80% in the presence of an NOS2 inhibitor, suggesting a fundamental role for NOS2 (Figure 3.4). In parallel, Nos2-/- MSC share a similar deficiency in suppressing splenocyte proliferation (Figure 3.12A). NO is considered to be a major mediator produced by NOS2, as accumulation of NO prevents Stat5 phosphorylation and in turn inhibits T cell proliferation [167, 448]. Of note, the supernatant collected from activated splenocytes was sufficient to induce up to 40μM NO production by MSC without cell-contact (Figure 3.13). Concomitantly, I observed no increase in the secretion of NO beyond basal levels by Nos2-/- MSC following stimulation.

As the anti-proliferative ability of MSC was not completely abolished, this might indicate the existence of other inhibitory mechanisms (Figure 3.4). Of note, the expression of Hdc was markedly induced by activated splenocytes in Nos2-/- MSC compared to WT MSC, suggesting that Hdc might contribute to the anti-proliferative activity found in Nos2-/- MSC (Figure 3.12C). However, the prominent role of NOS2 is likely to obscure the contribution of other molecules, making it difficult to assess their roles. In addition to the metabolic activities of HDC and NOS2 shown in my studies, the immunosuppressive activities of MSC have been demonstrated to be a concerted action of chemokines and NOS2 [79].

Much attention has been paid to studying the effect of histamine due to its dynamic role in many physiological conditions, including cell proliferation and differentiation, haematopoiesis, and wound healing [482-484]. Although conflicting results were obtained, these discrepancies might be illustrated by the differential expression of histamine receptors (H1, H2, H3, H4) by target cells. The H1 subtype is considered to be predominant in Th1 cells, whereas Th2 cells preferentially express H2 receptors [323]. Analysis of the effect of histamine on human Th1 and Th2 cells demonstrated that histamine promotes anti-CD3-induced proliferation of Th1 cells by triggering the H1 receptor, whereas the proliferation and the cytokine production of Th2 cells are inhibited. A controversial result has been reported from work on rodents, which showed that histamine shifts the Th1/Th2 balance towards Th2 by mediating the secretion of 117

Th1/Th2 cytokines [485]. However, the inhibitory effects modulated by histamine on antigen-induced murine T cell proliferation and IFN-γ production have been associated with both H1 and H2 receptors in vitro [486]. Here, my data suggested that HDC, the key catabolic enzyme for histamine synthesis, participates in MSC-mediated anti- proliferative activities (Figure 3.5). However, it remains to be identified whether this effect is based on uniform suppression or whether a particular subpopulation of splenocytes are suppressed. Interestingly, a recent study has indicated that histamine enhances IL-6 production by MSC through the H1 receptor, thereby promoting IL-6- dependent anti-apoptotic activity of MSC on neutrophils [487]. Therefore, histamine synthesised by HDC represents a possible target which may contribute to MSC- mediated immunomodulation via autocrine and paracrine regulation.

In contrast to NOS2, ARG inhibition did not reduce the ability of MSC to suppress lymphocyte proliferation (Figure 3.6). The role of ARG in competing with NOS2 and NOS3 for their common substrate, arginine, has been identified in MDSCs and endothelial cells, respectively [211, 488]. As specified, the advantage of ARG in driving T-cell suppression rests upon the absence of NO, when ARG antagonists effectively reverse MDSC-mediated inhibition [211]. Since both Arg and Nos2 transcripts are promoted by MSC, the effect of ARG may be transient as the dominant anti-proliferative effect is operated by NOS2 activity.

The anti-proliferative behaviour of MSC is not constitutive, but requires a licensing signal from the microenvironment. The pre-conditioning of MSC in order to enhance their immunosuppressive activity has an effect on the functioning of MSC. The expression of Arg1 can be induced by anti-inflammatory cytokines in macrophages and dendritic cells [188]. In agreement with this, my results indicate that both IL-4 and IL- 13 are effective in up-regulating Arg1 (Figure 3.8). LPS has been found to induce Arg1 in various tissues of rodents; this effect also occurred in the MSC [199, 489]. In contrast to Arg1, Arg2 is highly expressed following TNF-α stimulation. Since TNF-α is known to be released by activated splenocytes, it may provide an explanation for previously observed Arg2 induction in MSC cultivated with splenocytes (Figure 3.2).

Induction of Nos2 has been demonstrated to be induced by IFN-γ, TNF-α, IL-1, and LPS [79, 167, 490]. However, my results strongly suggested that, rather than IFN-γ, 118

TNF-α stimulation was a requirement for Nos2 up-regulation (Figure 3.8). The variation in cell isolation protocols, culture conditions, or selection of different markers may account for this difference. Although IFN-γ itself has a limited effect, it shows a synergistic effect in combination with TNF-α and it boosted the expression of Nos2 as previously reported [79, 491]. In addition, treatment with LPS also triggered the up- regulation of Nos2 in support of previous findings [79, 491].

TNF-α exerts its biological effect as an immunostimulant by binding to a 55-kd receptor (TNFR1) and a 75-kd receptor (TNFR2). My results indicated that ARG2 and NOS2 are both sensitive to the regulation exerted by TNF-α (Figure 3.8). Additionally, MSC generated from Tnfr1/r2-/- mice showed diminished expression of both Arg2 and Nos2 in response to activated splenocytes (Figure 3.14). In agreement with this, the protein levels of NOS2 and the production of NO were decreased by half in Tnfr1/r2-/- MSC compared to the control. Either neutralizing the TNF-α signal in the culture system or ablation of TNFR1/R2 in MSC led to a recovery of splenocyte proliferation, underlining the importance of this cytokine in MSC-mediated anti-proliferation.

My experiments on NF-κB, a master regulator of inflammation, further support the notion that the inflammatory environment is crucial for MSC licensing. While NOS2 production is mainly regulated at the level of transcription, the NF-κB binding site has been allocated to the region responsible for the induction of NOS2 by stimuli such as TNF-α, IFN-γ and LPS [492, 493]. As NF-κB has been shown to be a downstream target of the signalling cascades activated by TNF-α in MSC, it was not surprising that most of the inhibitors suppressing either IKKβ or IκB phosphorylation effectively abolished the expression of Nos2 (Figure 3.19). Further analysis, however, showed that both TPCA-1 and ML-120B failed to suppress the production of NO (Figure 3.20). It has been reported that IKK-independent pathways also provide mechanisms for integrating parallel signalling pathways that induce NF-κB activation. For instance, stimuli such as H2O2 also phosphorylate IκB, resulting in IκB degradation and activation of NF-κB [494]. It remains to be elucidated whether this mechanism could be responsible for sustained NF-κB activity in the presence of an IKKβ inhibitor. Nevertheless, it seems that Bay11-7085 represents a more prominent inhibitor by blocking NF-κB activity via directly targeting IκB phosphorylation. Although blocking NF-κB by Bay11-7085 has been observed to suppress the production of NOS2 in MSC, 119 the functional importance of this mechanism has not yet been demonstrated [490]. For the first time, my results indicate that loss of NF-κB activity abolishes MSC-mediated anti-proliferation, suggesting that the induction of NOS2 activity in MSC is NF-κB- dependent (Figure 3.21).

NF-κB signalling is also involved following TLR stimulation. I have observed that Nos2 up-regulation in MSC was enhanced by LPS (Figure 3.8). Ligation of LPS to TLR4 leads to the recruitment of adaptor proteins, such as MyD88 and Toll-interleukin 1 receptor domain-containing adapter protein (TIRAP) [495]. On the other hand, the binding of pro-inflammatory cytokine IL-1, to the IL-1R also results in the recruitment of MyD88 [57]. The activation of MyD88 triggers the NF-κB pathway and subsequently promotes the production of inflammatory cytokines. In the present study, I have demonstrated that Myd88-/- MSC were unable to generate Nos2 up-regulation in response to either LPS or IL-1β (Figure 3.23). Concordant with this observation, the inhibition of MyD88 by thalidomide has been shown to reduce NO production by impairing NF-κB activity in RAW 264.7 macrophage-like cells [496].

Nevertheless, Myd88-/- MSC displayed diminished, but not abolished, anti-proliferative ability, which was distinct from the incapacity observed in NF-κB-inhibited MSC (Figure 3.24). The remaining anti-proliferative activities in MSC could be explained by two mechanisms. Firstly, delayed activation of NF-κB has been observed in Myd88- /- cells, suggesting another adaptor protein, Trif, might be effective in promoting NF- κB with delayed kinetics [497]. Secondly, IRF3 has been shown to be responsible for the remaining NOS2 production in Myd88-/- macrophages [498]. Following LPS stimulation, IRF3 is phosphorylated and thereby promotes the transcription of IFN-β [499]. In turn, IFN-β binds to IFNR and activates JAK/STAT pathway, inducing transcription of Nos2 [500]. Unfortunately, the number of experiments conducted during my PhD was limited by the time required to generate MSC from KO mice – generally much longer than from WT mice. Therefore, confirmatory work and further experimentation are required to determine whether NOS2 or other factors are responsible for the persistent anti-proliferative activity in Myd88-/- MSC.

In conclusion, my results establish a clear profile of EAA-consuming enzymes, which provides an insight into the amino acid metabolism mediated by MSC during their anti- 120 proliferative activity. Additionally, I have demonstrated the importance of TNF-α as a licensing signal for NOS2 activity, which solely accounts for the anti-proliferative ability of MSC. While the mediators responsible for such regulation remain unclear, my studies have illustrated the involvement of NF-κB and MyD88 as up-stream mediators in regulating the suppressive effects of MSC (Figure 3.25).

Figure 3.25. EAA metabolic pathways regulated by MSC. In response to TNF-α, the activity of nitric oxide synthase 2 (NOS2) and arginase 2 (ARG2) is induced, which mediates the catabolism of arginine. Additionally, TNF-α and IFN-γ work synergistically to increase the production of NOS2. IFN-γ also regulate the activity of histamine decarboxylase (HDC), which controls the catabolism of histidine. Mechanistically, MyD88 and NF-κB are up-stream mediators of NOS2 activity.

121

4 THE EFFECT OF MSC ON HAEMATOPOIETIC CELLS: CELL CYCLE REGULATION AND APOPTOSIS

4.1 Introduction

It has been demonstrated that MSC exert their effect on immune cells through three aspects of regulation including: cell activation, proliferation and effector function. In the previous chapter, the involvement of EAA deprivation in the MSC mediated anti-proliferation effect was demonstrated. Having identified TNF-α-induced NOS2 activity as the main strategy for controlling T cell proliferation by MSC, dissecting this effect at the cell cycle level was a logical next step.

On the other hand, the ability of MSC to modulate the cell cycle of target cells may potentially contribute to their ability to maintain the HSPC compartment. HSPC comprises a rare population of cells residing in the BM, which undergo a highly co-ordinated differentiation and self-renewal program. A clear association between MSC and the HSC niche has been demonstrated by Mendez-Ferrer and colleagues by showing how Nestin+ MSC maintain HSPC in the bone marrow [398]. However, the precise mechanism underlying this regulation has yet to be elucidated. It has been demonstrated that HSPC actively interact with the microenvironment of BM controlling by cytokines, developmental signals, chemokines and adhesion molecules [353, 501, 502]. Maintenance of a quiescent state is the most critical feature for HSPC, as a loss of quiescence has been shown to impair the maintenance and function of HSPC [503]. Furthermore, HSPC have been shown to be highly sensitive to the metabolic regulation of glucose, which is capable of maintaining HSPC in quiescence [504]. Given the importance of nutrient limitation in haematopoiesis, and the fact that in the present study it has been demonstrated that MSC are capable of mediating metabolic pathways through EAA consuming enzymes, I sought to evaluate whether this mechanism is utilised by MSC to maintain HSPC dormancy.

122

4.2 Results

4.2.1. NOS2 plays a fundamental role in the MSC mediated cell-cycle arrest of activated splenocytes

Firstly, I sought to improve the characterisation of MSC-mediated cell-cycle regulation in splenocytes, in light of the involvement of TNF-α stimulation and NOS2 activity. Naïve or Con A-primed splenoyctes were cultured alone or co-cultured with WT, Tnfr1/r2-/-, and Nos2-/- MSC for 48 hours before their cell-cycle distribution was quantified by Hoechst 33342 and Pyronin Y staining. As shown in Figure 4.1, without

Con A stimulation, more than 90% of splenocytes were in the G0 phase (90.29±3.05%) and about 10% in the G1 phase (8.61±3.25%), whereas less than 1% of splenocytes were found in the G2/S phase (1.06±0.21%). In co-culture system containing both splenocytes and MSC, the cell-cycle of splenocytes was unaffected as the proportion of cells in each phase were similar to when splenocytes were cultured alone. Stimulation from Con A significantly decreased the proportion of splenocytes in the G0 phase from

90.29±3.05% to 38.90±4.23%, and increased the G1 fraction from 8.61±3.25% to

48.87±6.42%. Notably, WT MSC promoted the retention of splenocytes in the G0 phase (79.10±19.36%), while this effect was also observed with Tnfr1/r2-/- MSC (81.74±7.99%). In contrast, such a difference was not observed when splenocytes were -/- co-cultured with Nos2 MSC, when less than 50% of cells were retained in the G0 phase

(49.92±0.67%) and approximately 40% of cells were shifted into the G1 phase (38.78±1.96%). Collectively, these data suggest that MSC mediate splenocyte cell- cycle arrest through NOS2 activity, whereas TNF- α does not have a critical role in MSC regulation of the cell cycle.

123

) 100

(

y S/G2/M

o G

n 50 1

e l G

p 0

0 e n A C A C A C A o n S n S n S n l o M o - M o - M o A C C -/ C -/ C + T + + + W 2 r2 + C s C / C S o S r1 S M N - M f - M + -/ n -/ T 2 T 2 W s + /r + o 1 N fr + n T +

124

Figure 4.1. NOS2 deficiency, but not TNFR1/R2 deficiency, impairs the ability of MSC to induce the cell-cycle arrest of activated splenocytes. Con A-activated splenocytes were cultured with MSC isolated from WT, Tnfr1/r2-/-, Nos2-/- mice for 48 hours, followed by FACS analysis of the intercalation of Pyronin Y (DNA) and Hoechst 33342 (RNA). (A) Contour plots represent the CD45+ cells in the live gate. Numbers beside contour plots indicate the percentage of cells in each phase of the cell cycle. A representative example of two independent experiments is shown. (B)(C) Percentage of CD45+ cells in phases G0, G1 and S/G2/M of the cell cycle. Each bar indicates the mean ± SD from two independent experiments. * p < 0.05 ** p < 0.005; Unpaired t test.

125

4.2.2 MSC induced haematopoietic stem/progenitor cells (HSPC) cell-cycle arrest in vitro

After identifying NOS2 as a key molecule in inducing arrest in the G0 phase, I investigated whether MSC are capable of maintaining HSPC in a quiescent state through NOS2 activity. A combination of cytokines, including interleukin-3 (IL-3), stem cell factor (SCF) and Flt3- ligand (Flt3-L) were chosen to supplement HSPC maintenance in vitro [505, 506]. HSPC were co-cultured with WT or Nos2-/- MSC for 5 days, and their cell-cycle distribution was studied by assessing Hoechst 33342 and Pyronin Y staining (Figure 4.2). A difference in cell-cycle distribution was demonstrated when HSPC were co-cultured with MSC, although the MSC phenotypes did not account for the alterations observed (Figure 4.2A). -/- MSC cultivated with either WT or Nos2 promoted the ratio of HSPC in the G0 phase from 24±2.96% to 35.05±2.33% and 33.45±6.7%, respectively. In addition, the proportion of

HSPC in G1 was decreased to 48.65±3.04% in the presence of WT MSC, and to 49.30±6.50% with Nos2-/- MSC compared to 57.65±0.91% with HSPC.

To determine the influence of cellular proximity on HSPC cell cycle regulation, HSPC were divided into two subsets on the basis of their adherence to MSC. The non-adherent cells were collected from the supernatant and the remaining non-adherent cells were collected with gently washing of the MSC layer, while adherent cells were trypsinised and collected. The proportion of HSPC in suspension, which were retained in the G0 phase was progressively increased in the presence of WT MSC from 34.20±15.41% to 52.40±11.17%, as shown in Figure 4.2C, whereas the proportion of HSPC in the G0 phase in the presence of Nos2-/- MSC (29.70±3.95%) was similar to when HSPC were cultured alone. In parallel, fewer HSPC were observed in the G1 and replicating phase of the cell cycle (S/G2/M) after co-culture with WT MSC (54.95±16.75% versus 40.80±12.44% in the G1 phase,

9.62±1.95% versus 5.4±1.92% in the S/G2/M phase, in the absence versus in the presence of WT MSC). Nonetheless, the cell-cycle distribution of HSPC attached to the feeder layer were not identifiably different in the presence of WT MSC. Notably, when HSPC were -/- cultivated with Nos2 MSC, fewer cells were observed in the G0 phase (34.20±15.41% in the absence versus 30.95±0.07% in the presence of Nos2-/- MSC), whereas more HSPC were identified in the S/G2/M phase (9.62±1.95% in the absence versus 21.96±23.10% in the presence of Nos2-/- MSC).

126

127

Figure 4.2 MSC regulate the cell-cycle distribution of HSPC differently depending on cell contact. HSPC were maintained in FLT3-L (100ng/ml), SCF (100ng/ml), and IL-3 (10ng/ml) and cultured with MSC isolated from WT or Nos2-/- mice for 5 days, followed by FACS analysis of the intercalation of Pyronin Y (DNA) and Hoechst 33342 (RNA). HSPC were considered as one population in (A), or were separated into two populations depending on their localisation during co-culture in (B) (C). Contour plots represent the CD45+ cells in the live gate. Numbers beside contour plots indicate the percentage of cells in each phase of the cell cycle. A representative example of two independent experiments is shown. (A) + (C) Percentage of CD45 cells in phases G0, G1 and S/G2/M of the cell cycle. HSPC S, suspended HSPC; HSPC A, attached HSPC. Each bar indicates the mean ± SD from two independent experiments.

128

To dissect further the effect of MSC on the HSPC cell-cycle regulation, HSPC were maintained in culture medium without a cytokine supplement or were induced to enter the cell-cycle by G-CSF (Figure 4.3). In the absence of G-CSF stimulation, WT MSC induced the cell cycle entry of HSPC from G0 to G1 phase (19.55±8.15% versus

59.05±3.35% in the G1 phase, in the absence versus in the presence of WT MSC, p=0.0463) (Figure 4.3C). After G-CSF stimulation, the proportion of HSPC in the G0 phase dropped dramatically from 60.35±18.87% to 22.45±2.61%, whereas when they were cultivated with WT MSC, a similar ratio of HSPC were in the G0 phase (14.50±59.05% before versus 22.65±8.13% after G-CSF). The proportion of actively cycling HSPC (S/G2/M phase) was increased from 15.93±3.98% to 19.35±1.20% by G-CSF. Instead, the entry of HSPC to the S phase in response to G-CSF was suppressed by the presence of WT MSC (24.90±4.94% before versus 18.70±3.53% after G-CSF).

In order to evaluate the role of NOS2 in MSC mediated cell cycle arrest of HSPC, Nos2- /- MSC were co-cultivated with HSPC in the absence or presence of G-CSF. As observed in WT MSC, Nos2-/- MSC also promoted the cell cycling of HSPC in the absence of G-CSF (19.55±8.15% versus 57.80±0.50% in the G1 phase, in the absence versus in the presence of Nos2-/- MSC, p=0.0427) (Figure 4.3C). After G-CSF stimulation, the HSPC retained in the G0 phase were unchanged from 20.75±3.04% to 22±1.41%, whereas the proportion of cycling cells was slightly decreased from 19.95±1.90% to 17±1.55% in the presence of Nos2-/- MSC. The results suggest that, during the resting state, both WT and Nos2-/- MSC may promote the cell cycling of HSPC. Instead, following G-CSF stimulation, the presence of WT or Nos2-/- MSC seems to have a negligible effect on HSPC cell cycle regulation.

129

100

)

% S/G /M ( 2 C G P 50 1 S G H 0

0 e F C F C F n S S S S S o C C C l - M - /- M - A G T G - G + + 2 + W C s C + S o S N - M + -/ M T 2 W s + o N +

130

Figure 4.3. MSC maintain the cell-cycle distribution of HSPC in response to G-CSF stimulation. In the presence or absence of G-CSF (50ng/ml), HSPC cultured with MSC isolated from WT or Nos2-/- mice for 5 days, followed by FACS analysis of the intercalation of Pyronin Y (DNA) and Hoechst 33342 (RNA). (A) Contour plots represent the CD45+ cells in the live gate. Numbers beside contour plots indicate the percentage of cells in each phase of the cell cycle. A representative example of two + independent experiments is shown. (B)(C) Percentage of CD45 cells in phases G0, G1 and S/G2/M of the cell cycle. Each bar indicates the mean ± SD from two independent experiments. *p < 0.05; Unpaired t test. G-CSF, granulocyte-colony stimulating factor.

131

4.2.3 Effect of MSC on tumour cell proliferation

Another aspect investigated here was the interaction between MSC and tumour cells, in particular haematopoietic tumour cell lines. MSC were shown to provide crucial survival signals and to enhance the chemoresistance of tumour cells concomitant with decreased apoptosis [507-510]. On the other hand, several studies have demonstrated that the anti-proliferative effect of MSC resulted in significant inhibition of tumour growth, thereby promoting the survival of recipients [511-513]. Hence, the ability of MSC to inhibit tumour cell proliferation and the effect of MSC on drug-induced apoptosis was investigated here.

Firstly, the effect of MSC on malignant cell growth was evaluated. Tumour cell lines of haematopoietic origin, EL4 and MEL cells, were cultivated in different ratios with MSC. As shown in Figure 4.4, complete inhibition of EL4 proliferation was achieved in the presence of MSC at MSC:EL4 ratios of 1:4 to 1:16. The effect of MSC on suppressing MEL was less prominent as less than 35% inhibition was documented at the highest MSC:EL4 ratio tested (32.54±8.15% at MSC:EL4 ratio of 1:4). In accordance with data obtained from MSC/splenocyte co-cultures, MSC also exhibited a dose-dependent, anti-proliferative effect on both of the tumour cell lines investigated.

Since MSC potently suppress the proliferation of EL4, I further studied the effect of MSC on the etoposide-induced cell death of EL4. EL4 cells were cultivated alone or in the presence of MSC, the cells were harvested at the indicated time points following treatment with 1μM or 15μM etoposide (Figure 4.5A). The cell death was estimated by staining CD45+ EL4 cells with Annexin V (apoptotic cells) and 7-amino- actinomycin D (7AAD) (necrotic cells). After 2 hours, treatment with 1μM or 15μM etoposide induced a low degree of cell death of between 7% and 9%, whereas co- culturing with MSC reduced the cell death of EL4 to less than 5% (Figure 4.5B). The protective effect was evident after 8 hours, with the cell death of EL4 almost reduced to half in the presence of MSC (15.24% versus 6.66% and 22.74% versus 14.83%, EL4 versus EL4+MSC at 1μM and 15μM, respectively). After 16 hours, co-culturing with MSC decreased cell death induced by 1μM etoposide from 54.27% to 32.53%.

132

However, this protective effect of MSC was transient when the cells were cultured with 15μM etoposide; EL4 cell death was 88.71% in the presence of MSC as compared to 93.64% with EL4 alone.

100

)

( MSC+EL4

t MSC+MEL i 50

0 :4 :8 6 2 4 8 :4 :8 6 2 4 8 1 1 :1 :3 :6 2 1 1 :1 :3 :6 2 1 1 1 :1 1 1 1 :1 1 1

MSC:Tumour cells ratio

Figure 4.4. MSC exhibit dose-dependent anti-proliferative activity on tumour cell lines. EL4 and murine erythroleukemia (MEL) cells were plated at 5x104 cells/well in the presence or absence (controls) of C57BL/6 MSC at the indicated ratio for 3 days. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake, and the results are expressed as percentage of inhibition, obtained using the formula: {1-[(MSCTumourcells)- (MSC)/(Tumourcells)]}*100. Each column represents the mean ± SD of cultures in triplicate.

133

134

Figure 4.5. MSC is effective at preventing etoposide-induced apoptosis. 0.5x106 EL4 were cultured alone or with 0.125 x106 C57BL/6 MSC for 2, 8, or 16 hours. Cell death was induced by 1μM or 15μM etoposide and evaluated by FACS analysis according to 7AAD and Annexin V staining intensity. (A) Contour plots represent CD45+ cells in the live gate. Double negative cells were considered to be viable cells. 7AAD- Annexin V+ and 7AAD+ Annexin V+ cells were considered to be apoptotic cells. 7AAD+ Annexin V- cells were considered to be necrotic cells. (B) Percentage of EL4 cell death (with both apoptotic and necrotic cells included) at the indicated time points.

135

I then asked whether MSC could support the survival of EL4 under starving conditions (Figure 4.6). EL4 were cultured in serum-free conditions for up to 48 hours with or without WT MSC. At 12 hours, there was no difference observed between EL4 cultured alone (12.77%) and EL4 co-cultured with WT MSC (13.82%). A higher level of apoptosis was observed when EL4 were cultivated with WT MSC (20.15%) than in EL4 alone (15.68%) at 24 hours. Surprisingly, up to 92% of EL4 apoptosis was observed in the presence of WT MSC after 48 hours, whereas the proportion of apoptotic cells was only 33.60% in the absence of MSC. I investigated the effects of Nos2-/- MSC and Tnfr1/r2-/- MSC further in the same experimental setting. After 12 hours, the proportion of apoptotic EL4 cells was much higher when they were cultivated with Nos2-/- MSC (26.12% in the presence of Nos2-/- MSC versus 12.77% in the absence). In comparison, Tnfr1/r2-/- MSC appeared to reduce slightly the proportion of apoptotic EL4 cells from 12.77% to 9.82%. After 24 hours, EL4 had undergone double the amount of apoptosis when cultivated with Nos2-/- MSC (39.43%), whereas Tnfr1/r2- /- MSC did not affect the proportion of apoptotic EL4 cells (17.74%) when this was compared to EL4 cells cultured alone (15.68%). Like WT MSC, both Nos2-/- MSC and Tnfr1/r2-/- MSC greatly increased the apoptosis of EL4 from 33.60% to 92.39% and 82.62%, respectively. Thus, the results suggest that the protective effect of MSC on EL4 was only observed during etoposide-induced cell death. In contrast, MSC appears to be pro-apoptotic under conditions of serum starvation. However, these are only preliminary results and further studies will be required to determine the role of MSC in regulating the apoptosis of tumour cells under different circumstances further.

136

Figure 4.6. The effect of MSC on starvation-induced apoptosis. 0.5x106 EL4 were cultured alone or with 0.125 x106 WT, Nos2-/- or Tnfr1/r2-/- MSC in serum-free conditions for 12, 24, or 48 hours. Cell death was evaluated by FACS analysis according to the 7AAD and Annexin V staining intensity. (A) Contour plots represent CD45+ EL4 cells in the live gate. Double negative cells were considered to be viable. 7AAD- Annexin V+ and 7AAD+ Annexin V+ cells were considered to be apoptotic. 7AAD+ Annexin V- cells were considered as to be necrotic. (B) Percentage of EL4 apoptosis at the time points indicated.

137

4.3 Discussion

The inhibitory effect of MSC is particularly evident in NOS2-mediated T cell suppression. In parallel, the metabolism of arginine by either ARG or NOS is emerging as a fundamental mechanism for the regulation of immune responses [189, 514]. The resulting depletion of arginine has been associated with T-cell hyporesponsiveness in trauma [515], tuberculosis [516], pregnancy [517] and cancer [206]. Both in vitro and in vivo animal models have demonstrated that lack of arginine not only down-regulates T cell proliferation, but also induces molecular alterations including low expression of the CD3ζ chain and decreased cytokine production [205, 206, 518]. My study focused on the role of arginine in cell cycle regulation, and I demonstrated that the suppressed T cell proliferation mediated by MSC occurs via targeting the cell cycle. It has been suggested that depletion of arginine affects different stages through the cell cycle [519].

Cells can either be arrested in the G0/G1 phase, or subsequently enter the S phase. Otherwise, cells stop proliferating without completion of the cell cycle within the S phase. To characterise the cytostatic activity of NOS2 within the cell cycle, I performed an analysis to determine in which stage of the cell cycle the anti-proliferative effects take place. My results revealed a key feature of arginine depletion by showing that NOS2 activity in MSC not only prevented activated splenocyte entry into the S phase, but also prevented their entry into the G1 phase (Figure 4.1). As a consequence, the cells remained quiescent in the G0 phase. The effect was completely abolished by NOS2 deficiency, whereas both WT and Tnfr1/r2-/- MSC were capable of inducing cell cycle arrest. Interestingly, it seems that MSC only exhibited suppression of the cell cycle of activated splenocytes, but not of resting cells as stated in other studies [204]. Although the regulation of cell cycle mediators such as cyclin D2 and p27Kip1 were not investigated further in the present experiments, my results clearly demonstrated the profound effect of MSC on inducing cell arrest at G0 by NOS2.

Since NOS2 is a potent anti-proliferative factor, it may potentially contribute to HSC quiescence. In addition, the anti-proliferative activity of NOS2 via arginine depletion may overlap with its metabolic product, NO, which negatively regulates the cell cycle under certain circumstances. Based on the notion that MSC playing key roles in the niche and quiescence is a critical feature for HSC maintenance, I examined the role of

138

NOS2 expressed by MSC on maintaining the quiescence of HSPC in vitro. HSPC were stimulated by G-CSF to induce their proliferation in the presence of WT or Nos2-/- MSC. My results indicated that neither WT MSC nor Nos2-/- MSC exerted an effect on retaining HSPC in the G0 phase (Figure 4.3). In contrast, MSC induced the cell-cycling of HSPC in the absence of G-CSF.

NOS deficiency has been shown to increase the number of HSPC in bone marrow and the number of neutrophils in peripheral blood, suggesting their role is not redundant [520]. In vitro, HSPC cultured with MSC appeared to obtain an altered status when compared to HSPC alone. As suggested by others, the sensitivity to proliferative stimuli varies depending upon the differentiation status of the haematopoietic cells [521]. Therefore, supplemental cytokines, including FLT-3, SCF, and IL-3 were applied. No differences were observed by co-culturing the HSPC with either WT MSC or Nos2-/- MSC. Since cell contact with MSC has been shown to alter the functional, phenotypic, and clonogenic parameters of HSPC during co-culture, I analysed the HSPC as two subsets based on their adhesion to MSC [522]. It is plausible that this cultural system represents an environment where adherent HSPC constitutively supply the non- adherent HSPC with detaching cells. Interestingly, MSC markedly increased the proportion of non-adherent HSPC in the G0 phase, while the adherent HSPC shared a similar cell-cycle distribution with HSPC alone (Figure 4.2). Notably, this effect is absent from cultures with Nos2-/- MSC, suggesting a potential role for NOS2 in promoting the quiescence of HSPC. Other authors have also suggested that there is a spatial difference in the interaction between MSC and HSPC, where MSC retained more suspended HSPC in the G0/G1 phase, although the mechanism remains to be clarified [506].

In parallel with the fundamental role of MSC in HSC maintenance and expansion, MSC have been shown to protect haematopoietic malignant cells [523]. In addition, the protective effect of MSC is not limited to cells of haematopoietic origin, but is also active on cells of epithelial and nervous system origin [524]. It has been suggested that MSC contribute to tumour protection via their anti-proliferative effects, as they decrease the apoptotic activity of chemotherapeutic agents [510]. In the present study, I found that MSC were highly effective at suppressing EL4 cells over a wide range of

139

MSC/EL4 ratios (Figure 4.4). This was consistent with previous studies using other cell lines [525, 526]. However, I found that MSC were relatively less potent at suppressing MEL even at high ratios. Stromal cells were found to induce different mechanisms of drug resistance in target cells depending on the cell lines concerned, which may explain the differences observed in my study [527].

Further investigation revealed the effect of MSC on the survival of EL4 cells. The presence of MSC conferred protection from etoposide-induced necrosis of EL4 cells (Figure 4.5). Unexpectedly, the protective effect of MSC was transient at higher doses of etoposide, but this could be attributed to a cytotoxic effect on the MSC themselves. Previous studies have demonstrated that CML cells are protected by MSC from imatinib-induced cell death. MSC mediate this protective effect via SDF-1α secretion, which subsequently activates the CXCR4 pathway in CML cells and results in reduced caspase 3 activity [436]. In addition, MSC were also found to promote drug resistance in chronic lymphocytic leukaemia (CLL). Whilst circulating CLL cells are sensitive to forodesine-induced apoptosis, co-culturing CLL cells with MSC reduces the effectiveness of the drug and increases survival [528]. Together with these observations, my results support the notion that MSC are capable of preventing drug- induced apoptosis, and such effect is, at least in part, due to their anti-proliferative ability.

However, my results revealed a dual role for MSC under serum starvation (Figure 4.6). MSC not only displayed no protective effect on EL4, but also increased EL4 apoptosis. Interestingly, this apoptosis was enhanced when EL4 was cultivated with Nos2-/- MSC. Although these preliminary data remain to be confirmed, it should be mentioned that MSC have been observed to exhibit bidirectional effects on tumour progression. In particular, tumour necrosis factor-related apoptosis-inducing ligand (TRAIL) has been shown to be mediated by MSC to induce the apoptosis of tumour cells [529]. Furthermore, lentiviral transduced MSC expressing TRAIL were observed to reduce malignant pleural mesothelioma tumour growth both in vitro and in vivo via increasing tumour cell apoptosis [530]. Therefore, as plasticity is a fundamental feature of MSC,

140

MSC might have opposing effects on tumour growth depending on interactions between tumours, MSC, and the microenvironment. Further studies will be required to elucidate the role of MSC in tumour progression; apart from the relative number of MSC and tumour cells in culture, the tumour type, and the disease model should also be taken into consideration.

141

5 THE FUNCTION OF NOS IN THE MSC- MEDIATED IMMUNE RESPONSE DURING INFLAMMATION

5.1 Introduction

The immunomodulatory effect of MSC is influenced by the inflammatory milieu, whilst the resolution of inflammation involves fine adjustments of the cells involved in the systemic manifestation of the inflamed site. Thus, to confirm that the pathways I had identified in vitro also operate in vivo, I assessed the immunosuppressive abilities of MSC in two experimental models.

In the first system, an ex vivo model was applied to evaluate the effect of MSC on the prevention of antigen priming. Minor histocompatibility (H) antigens constitute a barrier to transplantation between individuals who have been matched for the dominant antigens encoded by MHC class I and II molecules [531]. They are peptides, which induce cell-mediated responses when recognised by allogenetic T cells. Among these, HY peptides are the male-specific, Y chromosome-encoded minor H antigens [532]. Minor H antigens, including HY antigens, have been linked to the development of GvHD after bone marrow grafts [533]. As the peptide identity and the MHC restriction of the immunodominant epitopes have been clearly defined, the immune response against HY peptide was chosen to study the immunosuppressive effect of MSC in vivo [158, 534, 535]. The two MHC class I-restricted peptides applied in the study, WMHHNMLDI and KCSRNRQYL, originated from the Uty and Smcy genes, respectively. In addition, WT female B6 (H2b) strains were applied as recipients since as they have been reported to be strong responders to HY-disparate skin grafts [536].

The second model applied to recapitulate an innate immunity reaction was TG-induced peritonitis. The TG model was first introduced as a method for generating macrophages

142

[537]. Non-enzymatic reactions between proteins and reducing sugars in TG broth result in the formation of advanced glycation end-products[538]. These end-products are recognised by members of the PRR family, including macrophage scavenger receptors and advanced glycation end-product receptors [539]. Upon inflammation, rapid infiltration and accumulation of neutrophils in the peritoneal cavity occurs within 12 to 24 hours [540] (Figure 5.1). A more sustained population composed of monocytes and/or macrophages subsequently replaces the neutrophils [541]. The final phase is characterised by prominent lymphocyte recruitment, which repopulates the cavity upon the resolution of inflammation [542]. The use of TG-induced peritonitis offers several advantages, the first of which is the opportunity to investigate inflammation in a confined space such as the peritoneal cavity. This model also provides a well-characterised platform for studying the effects of MSC on leukocyte recruitment during acute inflammatory conditions.

Figure 5.1. Leukocyte kinetics during TG-induced peritoneal inflammation. Schematic representation derived from data in murine models of acute inflammation. Adapted from [543].

143

5.2 Results

5.2.1 HY responses in mice immunised with male splenocytes

WT female B6 (H2b) mice were immunised by intra-peritoneal (i.p.) injection of 15x106 syngeneic male splenocytes at day 0 and day 7. Fourteen days is the usual time required following immunisation to obtain HY-specific CD8+ T cell expansion in recipients [534]. Therefore, at day 14, spleen cells from immunised mice were harvested and restimulated with H2b-restricted HY peptide-pulsed (HY DbSmcy and HY DbUty) female splenocytes. The T-cell proliferation induced was assessed by incorporation of [3H] thymidine (Figure 5.2). The ex vivo stimulation generated a compatible level of proliferation in response to graded doses of HY peptide pulsing of HY-specific memory cells: HY-pulsed cells in a ratio of 1:2. Nonetheless, pulsing with doses of HY peptides as high as 100μg/ml did not induce much of a difference in proliferation when compared to doses of only 10μg/ml (14165.67±3151.62 versus 14573.33±2037.91 with Uty, 14807.33±2989.48 versus 14495.33±1453.49 with Smcy, 10μg/ml versus 100μg/ml at a ratio of 1:2, respectively). In addition, both peptides induced the proliferation of HY-specific memory T cells to a similar level between 10000 to 15000 cpm, whereas without peptide restimulation, the background radiation was less than 5000 cpm.

Next, I evaluated the immunosuppressive ability of MSC in vivo by comparing HY-specific memory T cell generation in the presence or absence of MSC. Female mice were immunised with 15x106 syngeneic male splenocytes at day 0, day 7 and day 14, whilst 15x106 MSC were administered at day 7 and day 14. At day 21, the cells were harvested following restimulation with H2b-restricted HY peptide-pulsed (HY DbSmcy and HY DbUty) female splenocytes or syngenic male splenocytes. As shown in Figure 5.3, the background response of splenocytes alone was as high as the proliferation stimulated by HY peptides, leading to unjustified results. Also, the proliferation levels between individual responder receiving MSC varied within experimental groups. Repeated experiments still resulted in high background threshold discriminates, making the data obtained redundant.

144

Figure 5.2. Quantification of HY-primed T cells stimulated in vivo. Female C57BL/6 mice were immunised with 15x106 male splenocytes at day 0 and day 7. Mice were sacrificed after 14 days and the splenocytes were harvested and plated at 0.5x106 cells/well. Irradiated (3000 rads) female splenocytes as APC (2x106, 1x106, 0.5x106, and 0.25 x106) were labelled with H-Y peptide (A) Uty (B) Smcy in various concentrations (0.1-100μg/ml) for 1 hour and cultured with primed splenocytes for 72 hours. The bottom panel indicates the ratio between primed splenocytes against APC. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake. The figure illustrates data from a representative experiment, and each column represents the mean ± SD of cultures in triplicates. cpm, counts per minute.

145

Figure 5.3. Effect of MSC on HY priming in vivo. Female C57BL/6 mice were immunised with 15x106 male splenocytes at day 0, day 7 and day 14. MSC (15x106) were injected at day 7 and day 14. Mice were sacrificed after 21 days and the splenocytes were harvested and plated at 0.5x106 cells/well. Irradiated (3000 rads) female splenocytes as APC (1x106) were labelled with H-Y peptide (A) Uty (B) Smcy at various concentrations (0.1-10μg/ml) for 1 hour and cultured with primed splenocytes for 72 hours. (C) Irradiated (3000 rads) male splenocytes as APC (2x106, 1x106, 0.5x106, 0.25x106) were cultured with primed splenocytes for 72 hours. 0.5 μCi/well of [3H] thymidine was added for 18 hours prior to the harvest of cells. Proliferation was determined as a measure of [3H] thymidine uptake. Each column represents the mean ± SD of data from cultures in triplicate. Spl, mice injected with male splenocytes. Spl+MSC, mice injected with male splenocytes and MSC. cpm, counts per minute.

146

5.2.2 MSC markedly reduced the TG-induced inflammatory infiltration

B6 mice were injected (i.p.) with TG or with PBS as control. Thirty minutes later, mice received through the same route 2 x 106 WT MSC or PBS as control, and were sacrificed at the specified time points (Figure 5.4A). After 18 hours, 24 hours, 3 days and 6 days, the cells in the peritoneal cavity were harvested and counted (Figure 5.4A). Injection of TG (i.p.) induced a prompt and acute inflammatory response in the mice, as indicated by the increase in the total number of cells in the peritoneal lavage. An 11- fold increase was observed after 18 hours, 8-fold after 24 hours, and a steady decline thereafter, with all results obtained by comparing to unstimulated mice (approximately 3x106 peritoneal cells). Administration of MSC after TG injection resulted in a 72% decrease in the total number of peritoneal cells by 18 hours, more than 60% by 3 day, and 32% by 6 days. Injection of PBS or of MSC did not produce any significant change in the total number of peritoneal cells. Due to these data, day 3 after TG treatment was selected to be the optimum time to address the role of MSC in immune cell recruitment.

Figure 5.4. Infiltrated peritoneal cells in response to TG-induced peritonitis. TG, or PBS (control) was administered intraperitoneally 30 minutes before MSC administration (2x106 cells/mouse). The total number of cells in the peritoneal cavity was assessed at different time points: 18 h, 24 h, 3 days and 6 days after TG injection. MSC injection decreased the inflammatory response to TG-induced peritonitis. Data represent means ± S.D. and are pooled from at least 3 independent experiments per time point (n ≥ 10 mice per time point). *** p < 0.0005; **** p < 0.0001; Unpaired t test.

147

5.2.3. Nos2-/-MSC regulate leukocyte infiltration differently during inflammation

To evaluate the role of NOS2 as an immunosuppressive factor in peritoneal inflammation, 2 x 106 Nos2-/- MSC or WT MSC were injected into B6 recipients following the i.p. administration of TG. Surprisingly, WT and Nos2-/- both exerted an immunosuppressive effect and suppressed cell recruitment in the peritoneal lavage. The administration of WT MSC and Nos2-/- MSC produced a 50-fold and 60-fold decrease, when compared to the TG group (approximately 20 x 106 peritoneal cells), respectively (Figure 5.5A). To characterise the impact of MSC on leukocyte recruitment further, the accumulated cell population was grouped on the basis of their surface marker expression (Figure 5.5B).

For this purpose, CD11b (Mac-1) expression was applied as a general marker for myeloid cells. The marker, Gr-1, is a composite epitope that recognises both Ly6G, which is expressed principally on mouse neutrophils [544], and Ly6C, which is broadly expressed by various haematopoietic cells, including T cells and monocytes [545, 546]. Therefore, neutrophils were characterised by their expression of CD11b, Ly6G, and Ly6C (CD11b+Ly6G+Ly6C+cells) [547]. Monocytes were subdivided into two subsets according to their expression of Ly6C and CCR2 (MCP-1); CD11b+Ly6C-CCR2- cells as resident monocytes, CD11b+Ly6C+CCR2+ cells as inflammatory monocytes [548]. Macrophages were identified by their expression of CD11b, F4/80, and lack of Gr-1 (CD11b+F4/80+Gr-1- cells) expression [549].

As shown in Figure 5.5C, the presence of both WT and Nos2-/- MSC significantly decreased the accumulation of CD11b+Ly6G+Ly6C+cells triggered by TG (0.62x105±0.30x105 cells with WT MSC +TG, 0.49x105±0.30x105 cells with Nos2-/- MSC+TG, versus 1.4x105±0.6x105 cells with TG, p=0.035 and 0.025, respectively). A significant decrease was observed in the absolute number of CD11b+Ly6C-CCR2- cells with WT MSC administration (19.07x105±9.66x105 cells versus 45.22x105±16.66x105 cells, p=0.0208, WT MSC+TG versus TG). Both types of MSC treatment resulted in a decrease in the total number of CD11b+Ly6C+CCR2+ cells (0.73x105±0.23x105 cells with WT MSC+TG, 0.60x105±0.32x105 cells with Nos2-/- MSC+TG, compared to 1.32x105±0.52x105 cells with TG). Analysis of the CD11b+F4/80+Gr-1- cell numbers

148 revealed a noticeable, but not statistically significant difference, between each group. With the administration of WT MSC, the total number of CD11b+F4/80+Gr-1- cells induced by TG dropped from 53.96x105±21.01x105 to 19.18x105±15.25x105, whereas Nos2-/- MSC treatment produced a less prominent decrease (26.95x105±18.96x105 cells).

The percentages of four cell populations corresponding to the total peritoneal cell population are presented in Figure 5.5D. A pattern similar to absolute numbers was observed in the percentage of CD11b+Ly6G+Ly6C+cells. Apart from the high variation present in the WT+TG group, the treatment of Nos2-/- MSC decreased the percentage of CD11b+Ly6G+Ly6C+ cells approximately by half (0.56%±0.14% versus 0.87%±0.34%, Nos2-/- MSC+TG versus TG group). As observed with the absolute number of CD11b+Ly6C-CCR2- cells, the percentage of cells was noticeably decreased by WT MSC administration (22.02%±9.25%), whereas Nos2-/- MSC did not produce a difference (28.48%±10.71% versus 27.90%±8.91%, Nos2-/- MSC+TG versus TG group). Similar proportions of CD11b+Ly6C+CCR2+ cells were present in all groups (approximately 1%). Following the trend documented in the total number of CD11b+F4/80+Gr-1- cells, the administration of WT MSC and Nos2-/- MSC generated a distinguishable, albeit low reduction in the percentage of CD11b+F4/80+Gr-1- cells (27.38%±12.41% in the WT MSC+TG group, 30.74%±10.65% in the Nos2-/- MSC+TG group, versus 33.32%±11.53% in the TG group). Collectively, WT MSC and Nos2-/- MSC were both effective in suppressing TG-induced peritoneal inflammation; in addition, they exhibited different impact on the regulation of the recruitment of distinct leukocytes.

149

A ) 6 0

1 4 0 x (

s l l

e 3 0 c T G d e

t T G + W T M S C a

r 2 0

t * * * * * -/- l

i T G + N o s 2 M S C f n i

P B S l 1 0 a e n o t i

r 0 e P

TG WT MSC+TG Nos2-/- MSC+TG PBS

Ly6C

CD11b Ly6C

F4/80

150

Figure 5.5. NOS2 had limited impact on the immunosuppressive effect of MSC in TG-induced peritonitis. TG or PBS was administered intraperitoneally 30 minutes before MSC administration (2x106 cells/mouse). Mice were sacrificed 72 hours after TG injection. (A) Total cell migration to the peritoneal cavity was counted using a Neubauer chamber. Each symbol represents an individual mouse; longer horizontal lines indicate the mean; shorter horizontal lines indicate the S.D. from two independent experiments (n=5). (B) Flow cytometric characterisation of CD11b+Ly6G+Ly6C+ cells (neutrophils), CD11b+Ly6C-CCR2- cells (resident monocytes), CD11b+Ly6C+CCR2+ cells (inflammatory monocytes), CD11b+F4/80+Gr1- cells (macrophages) in the peritoneal cavity. (C) Number, or (D) percentage of CD11b+Ly6G+Ly6C+ cells, CD11b+Ly6C-CCR2- cells, CD11b+Ly6C+CCR2+ cells, and CD11b+F4/80+Gr1- cells in the peritoneal cavity. Data from one representative experiment out of two independent experiments is shown. Data represent means ± S.D. from at least 5 animals per group. *p < 0.05 **p < 0.005 ***p < 0.001; Unpaired t test.

151

5.2.4. TNF-α plays a role in suppressing the percentage of leukocyte infiltrate during inflammation

Licensing by TNF-α represents the main route for stimulating the anti-proliferative activity of MSC in vitro. Therefore, I evaluated if withdrawing the effect of TNF-α would influence the immunosuppressive activity of MSC during peritoneal inflammation. The administration of WT and Tnfr1/r2-/- MSC both resulted in a 50% of reduction in the total number of peritoneal cells compared to the TG group (Figure 5.6A). Following the strategy for characterisation previously described, four subpopulations of leukocytes were assessed: neutrophils (CD11b+Ly6G+Ly6C+cells), resident monocytes (CD11b+Ly6C-CCR2-cells), inflammatory monocytes (CD11b+Ly6C+CCR2+cells) and macrophages (CD11b+F4/80+Gr-1-cells) (Figure 5.6B).

As previously described, following the administration of WT MSC, the influx of CD11b+Ly6G+Ly6C+cells triggered by TG was significantly reduced by half (0.82x105±0.33x105 cells versus 2.03x105±1.22x105 cells, p=0.0429, WT+TG group versus TG group) (Figure 5.6C). The overall suppressive effect was not different in animals treated with from Tnfr1/r2-/- MSC (0.99 x105±0.17x105 cells). However, differences were present in the different populations of accumulating immune cells. In comparison with the marked increase in CD11b+Ly6C-CCR2-cells after TG stimulation, both WT and Tnfr1/r2-/- MSC inhibited cell accumulation in the peritoneal cavity (36.92x105±14.17x105 cells with WT MSC+TG, 34.81x105±6.30x105 cells with Tnfr1/r2-/- MSC+TG, versus 79.72x105±15.73x105 cells with TG, p=0.001 and 0.0004, respectively). The presence of WT MSC resulted in a noticeable decrease in CD11b+Ly6C+CCR2+cells when compared to the TG treated group, whereas this effect was not observed after the treatment with Tnfr1/r2-/- MSC (1.21x105±0.39x105 cells with WT MSC+TG, 2.39x105±1.40x105 cells with Tnfr1/r2-/- MSC+TG, versus 2.84x105±1.82x105 cells with TG). Following a TG stimulus, accumulation of CD11b+F4/80+Gr-1-cells in the peritoneal cavity was high. This increase was significantly diminished by the presence of both WT and Tnfr1/r2-/- MSC

152

(57.66x105±22.61x105 cells with WT MSC+TG, 59.94x105±7.03x105 cells with Tnfr1/r2-/- MSC+TG, versus 120.26x105±15.73x105 cells with TG, p<0.0001).

When the infiltrating cells were quantified as a percentage of the total peritoneal cell population, the pattern of the results was slightly different. Only WT MSC documented a marked decrease in CD11b+Ly6G+Ly6C+cells after TG stimulation (0.56%±0.15% versus 0.96%±0.41%, WT MSC+TG versus TG) (Figure 5.6D). Both WT MSC and Tnfr1/r2-/- MSC were effective in reducing the proportion of CD11b+Ly6C-CCR2-cells in the peritoneal cavity in response to TG (24.65%±5.01% versus 39.52%±5.97%, p=0.0015, and 29.08%±2.55% versus 39.52%±5.97%, p=0.0071, WT MSC+TG versus TG, and Tnfr1/r2-/- MSC versus TG, respectively). In the presence of WT MSC, the percentage of CD11b+Ly6C+CCR2+cells decreased markedly from 1.33%±0.66% to 0.87%±0.41% as compared to the TG treated group. In contrast to the results derived from the absolute cell numbers, the percentages of CD11b+F4/80+Gr-1-cells led to the suggestion that WT MSC were more potent at reducing the proportion of macrophages as compared to Tnfr1/r2-/- MSC (45.60%±5.62% versus 54.34%±6.94%, p=0.0462, and 50.94%±7.51% versus 54.34%±6.94%, WT MSC+TG versus TG, and Tnfr1/r2-/- MSC versus TG, respectively). Although WT and Tnfr1/r2-/- MSC were both effective in suppressing TG-induced leukocyte influx, the results here suggested that WT MSC were more effective at reducing the proportion of leukocyte subpopulations.

153

A ) 6 0

1 4 0

x (

s l l

e 3 0 T G c

d * * * e * * * T G + W T M S C t

a 2 0 r -/- t

l T G + T n fr 1 /r 2 M S C i f n

i P B S

l 1 0 a e n o t i

r 0 e P

TG WT MSC+TG Tnfr1/r2-/- MSC+TG PBS

Ly6C

CD11b Ly6C

F4/80

154

Figure 5.6. Deficiency of TNFR1/R2 modulates the immunosuppressive effect of MSC in TG- induced peritonitis. TG or PBS was administered intraperitoneally 30 minutes before MSC administration (2x106 cells/mouse). Mice were sacrificed 72 hours after TG injection. (A) Total cell migration to the peritoneal cavity was counted using a Neubauer chamber. Each symbol represents an individual mouse. Longer horizontal lines and shorter horizontal lines indicate the mean and S.D., respectively, of two independent experiments (n=5). (B) Flow cytometric characterisation of CD11b+Ly6G+Ly6C+ cells (neutrophils), CD11b+Ly6C-CCR2- cells (resident monocytes), CD11b+Ly6C+CCR2+ cells (inflammatory monocytes) and CD11b+F4/80+Gr1- cells (macrophages) in the peritoneal cavity. (C) Number or (D) percentage of CD11b+Ly6G+Ly6C+ cells, CD11b+Ly6C-CCR2- cells, CD11b+Ly6C+CCR2+ cells and CD11b+F4/80+Gr1- cells in the peritoneal cavity. Data from one representative experiment of two independent experiments is shown. Means ± S.D. were calculated from at least 5 animals per group. *p < 0.05 **p < 0.005 ***p < 0.001 ****p < 0.0001; Unpaired t test.

155

5.3 Discussion

In line with the promising results on the immunoregulatory activities of MSC in vitro, animal models have also demonstrated that MSC treatment can be beneficial in a diverse range of disease models. For instance, MSC have been shown to reduce inflammation in including colitis [127, 550], rheumatoid arthritis [551, 552], myocarditis [553], and GvHD [554, 555].

Previous studies have shown that the administration of MSC in baboons significantly prolong the survival of MHC-mismatched skin grafts [160]. In addition, our research group has previously demonstrated that MSC have a profound inhibitory effect on the HY-specific proliferation of both naïve and memory T cells in vitro [158]. To examine directly the immunosuppressive features of MSC in inhibiting antigen-reactive T cells in vivo, I established an ex vivo system for detecting the priming of T cells against HY.

I did not detect any effect of MSC on the generation of HY-specific memory T cells (Figure 5.3). However, there are a number of factors that are intrinsic to the model that need to be taken into account. The proliferation of HY-specific T cells induced in the HY-primed female splenocytes was not very reproducible. To enhance the difference in the proliferation of the T cell population of interest, a second re-stimulation may be required to produce a dominant expansion of HY-specific T cells [534].

Another aspect to consider is the time taken for MSC licensing. Based on the results from previous chapters and other studies, the inflammatory factors secreted during the immune response are fundamental to turn on the immunosuppressive capacity of MSC. Although it has been reported that CD8+ T cells in B6 female recipient mice generated a panoply of cytokines and chemokines, including TNF-α, XCL1, CCL3, and CCL4, following immunisation with HY peptide, these are not easily traceable [556]. Therefore, as the cytokines and chemokines are only produced for a short time frame in response to HY, the concentrations of ‘licensing signals’ may not be sufficient to activate MSC. The signals could also have been lost before the MSC arrived at the inflammatory site in my studies. In fact, the timing of MSC administrations is another factor that is critical to their functions. Two different studies on the effects of MSC on

156 preventing GvHD in mice have led to distinct results [79, 418]. A possible explanation for this difference is the infusion time of MSC; whether the timing of MSC administration corresponds with the peak of inflammation appears to affect their therapeutic effects.

Further factors that might have impacted on the experimental setting is the ability of MSC to traffic to the site of HY-specific inflammatory effectors. Based on the results documented in previous chapters, MSC exhibit their immunomodulatory capacity via NO. Therefore, it is conceivable that MSC may only inhibit target cells that are in close proximity, while my data might reflect the inability of MSC to reach the inflammatory site at which immunisation took place. Guided by this presumption, I have then analysed the immunomodulatory activity of MSC in another experimental model: TG- induced peritonitis. It has been demonstrated that TG administration induces acute inflammatory responses accompanied by pro-inflammatory cytokine production and leukocyte accumulation [557]. Thus, it provides an ideal microenvironment to investigate the effects of licensing signals on the function of MSC. In addition, as MSC are kept in close contact with immune cells recruited to the peritoneum during the course of inflammation, it is an appropriate model to elucidate how NOS2 and its product NO is used as a mediator of immune regulation by MSC.

MSC treatment remarkably reduced the local inflammatory response during acute peritoneal inflammation assessed by the total cell numbers mobilised into the peritoneal cavity (Figure 5.4). The migration of leukocytes is regulated by the gradients of chemoattractant in the peritoneum, which undergoes a fast increase following thioglycollate treatment [557]. Since MSC generated an overall suppression of inflammatory cell recruitment, this result can be interpreted as the suppressive effect of MSC on chemokine production, either directly or indirectly via candidate cells such as macrophages. In fact, MSC-mediated chemokine production has been reported in a mouse model of zymosan-induced peritonitis [104]. Firstly, MSC were observed to suppress the production of pro-inflammatory cytokines by zymosan-activated resident macrophages by secreting TSG-6. As a consequence, the release of IL-6 and CXCL1, which amplify the pro-inflammatory signals, by mesothelial cells was interrupted and the recruitment of neutrophils was diminished. As shown in Figure 5.5, MSC potently attenuated neutrophil infiltration, as well as the influx of resident monocytes, 157 inflammatory monocytes, and macrophages. Considering the role of monocytes/macrophages in the clearance of neutrophils by phagocytosis, the reduced number of apoptotic neutrophils due to lower neutrophil accumulation might subsequently decrease recruitment of monocytes/macrophages. Collectively, the TG model provided an ideal system for testing the role of not only the effector molecules, but also those of the licensing pathways identified in vitro.

Surprisingly, regardless of the diminished anti-proliferative activity of Nos2-/- MSC in vitro, they were as effective as WT MSC at reducing the total number of peritoneal cells. Nevertheless, NOS2 deficiency diminished the inhibitory effect of MSC on suppressing macrophage infiltration (Figure 5.5). Since macrophages are important effector cells in both innate and adaptive immune responses, it is of interest to pinpoint the role of NOS2 in MSC-mediated macrophage recruitment in future studies. Although in the mouse model of peritonitis, MSC exhibit anti-inflammatory effects without the presence of NOS2 induction, that in other inflammatory diseases, NOS2 may be a contributing factor exerted by MSC cannot be excluded. In fact, the induction of NOS2 has been shown to play a critical protective role in acetic acid-induced acute colitis by reducing leukocyte infiltration [558].

On the other hand, administration of Tnfr1/r2-/- MSC produced a similarly inhibitory effect to WT MSC. In addition to the suppressed cell numbers that accumulated in the peritoneal cavity, Tnfr1/r2-/- MSC regulated the recruitment of neutrophils, resident/inflammatory monocytes, and macrophages in a similar pattern to WT MSC (Figure 5.6). When the population of inflammatory cells was presented as a relative percentage of the total cell number, the suppressive effect was absent, suggesting that MSC induced a general reduction in the infiltrating cells independent of cell type (Figure 5.6).

Although TNF-α-mediated NOS2 production seems to be a fundamental mechanism contributing to the anti-proliferative activity of MSC in vitro, my results suggest that this mechanism might be complemented by other processes which also deliver the immunosuppressive effect of MSC in vivo. For instance, MSC can modify the inflammatory environment by reducing levels of TNF-α and IL-6, and increasing levels of IL-4 and IL-13 following transplantation. This effect involves a shift in the 158 macrophage phenotype from M1 to M2, and has been reported in acute inflammation during spinal cord injury, while the administration of MSC attenuates inflammation and promotes tissue repair [156]. In addition, MSC production of TGF-β and PGE2 was found to promote Treg expansion, and subsequently ameliorate allergic airway inflammation and improve lung function [559]. Interestingly, as MSC favour the differentiation of M2 macrophages, the subsequent production of IL-10 and CCL18 by M2 macrophages also promotes Treg induction [101]. Therefore, the fact that MSC regulate the proliferation and differentiation of immune cells, along with the concept that MSC modulate the chemokine and cytokine profiles of the inflammatory environment, may substitute for the effects of TNF-α-mediated NOS2 induction and produce the immunomodulatory activities of MSC.

159

6 GENERAL DISCUSSION

My project has addressed basic issues in MSC immunobiology that have important implications in MSC therapeutics. Despite over 300 trials employing MSC for different conditions, therapeutic efficacy of MSC has not always been achieved for all measured outputs. My results have clearly indicated the importance of MSC being exposed to the appropriate inflammatory environment to acquire immunosuppressive functions. The signals provided by the niche have been identified in vitro but their impact in vivo has not been confirmed. For example, the potent anti-inflammatory effect that we observed in a peritonitis model was still delivered from MSC with the NOS enzyme knocked- out. This suggests that other factors can replace NOS in vivo. Therefore, to better understand the mechanisms responsible for MSC-based therapy, the dynamic nature of the tissue microenvironment should also be taken into consideration in order to validate in vitro results.

The first factor to discuss is MSC trafficking during inflammation. MSC have a tendency to home towards inflammatory sites. It is presumed that inflammation- directed MSC homing is required for their therapeutic efficacy due to exposure to licensing signals. Several cell trafficking-related molecules have been associated with MSC homing activity, such as chemokines, adhesion molecules and matrix metalloproteinases (MMPs). The CXCL12/CXCR4 axis plays an important role in mediating the recruitment of MSC to injury sites such as bone fractures, or myocardial infarctions. Furthermore, binding to the endothelial lining via adhesion molecules, like P-selectin and the VCAM-1/ very late antigen-4 (VLA-4), is fundamental for MSC adhesion and rolling within inflamed tissues [560]. The control of MSC trafficking involves their invasion through barriers of the ECM. It has been shown that the invasive activity of MSC is substantially dependent on the production of MMP-2 and membrane type 1 MMP (MT1-MMP) [561]. It is of note that the induction of these homing-related molecules can be modified by inflammatory cytokines, including IL-1β, TNF-α, and TGF-β1. These findings suggest the need to identify how different inflammatory profiles lead to distinct MSC trafficking, engraftment and therapeutic activities.

160

The second aspect concerns the MSC response to local inflammatory stimuli. These are indispensable for stimulating MSC to secret factors that are critical in suppressing inflammatory responses and mediating tissue preservation. Thus, it is conceivable that MSC administration at the peak of an inflammatory response may improve their therapeutic efficacy. In an effort to determine the optimal schedule for MSC infusion, a series of experiments has been performed in a murine GvHD model. No significant efficacy was observed when MSC were given either at the time of transplant or on day 30, while MSC significantly improved GvHD mortality when given during ongoing GvHD on day 20 [420].

MSC were found to modify their transcriptome for anti-inflammatory and anti- apoptotic factors in response to the ischemic microenvironment within 24 hours from their administration [562]. In parallel, more than 10% of ischemia-induced genes were found to be down-regulated. This highlights that the therapeutic effect of MSC is mediated by the ischemic microenvironment or by cross-talk with cells in the vicinity. During bleomycin-induced lung inflammation, which is an IL-1 dominant microenvironment, MSC were observed to efficiently localise in the inflamed lung and subsequently produce a significant amount of interleukin 1 receptor antagonist (IL1RN) [563]. Interestingly, MSC functioned more effectively than recombinant IL1RN in inhibiting immune cell recruitment and inflammatory cytokine production. This could be explained by the fact that MSC function both directly and indirectly: directly by impairing the effectors of inflammation and indirectly by recruiting and activating other cell types with immunomodulatory effects.

MSC regulate the recruitment of leukocytes during inflammation through an interaction with neighbouring endothelial cells. Accordingly, MSC facilitate the establishment of tissue ‘address-codes’ that direct leukocyte behaviour in the inflamed tissue [564]. Coordinated interactions among stroma, endothelial cells and leukocytes dictate which leukocytes are recruited, whether they are retained within the inflamed site and their life span [565]. In response to TNF-α stimulation, MSC were observed to suppress neutrophil adhesion to endothelial cells during co-culture [566]. Also, cross-talk between MSC and endothelial cells up-regulated the production of IL-6 by MSC, which

161 in turn decreased the response of endothelial cells to inflammatory cytokines [567], so that their ability to recruit leukocytes was reduced.

In addition to preventing the recruitment of leukocytes, MSC also contribute to accelerating the resolution of inflammation via macrophages. MSC have been shown to suppress the production of the inflammatory cytokines, TNF-α, IL-6, IL-12p70 and IFN-γ, in vitro, but to promote the production of IL-10 and IL-12p40 by thioglycolate- elicited peritoneal macrophages after LPS stimulation [568]. Such an effect is applied by MSC to suppress inflammation at inflamed sites, where pro-inflammatory cytokines are abundantly secreted by monocytes and macrophages. Furthermore, this results in an impaired induction of Th1 responses. Th1 responses.

The inhibition of pro-inflammatory mediators by MSC has been correlated with the phagocytosis of apoptotic neutrophils by macrophages. Macrophages are thought to gradually lose their ability to produce pro-inflammatory cytokines after taking up apoptotic cells during inflammation [557]. Therefore, MSC can attenuate the inflammatory cascade by switching macrophages to a regulatory phenotype that is also characterised by an increased ability to phagocytose apoptotic cells. There is evidence that macrophages produce high levels of IDO following the phagocytosis of apoptotic cells [569]. The therapies mediated by MSC via macrophage regulation have been demonstrated in ischemia and different models of acute inflammation, such as cecal ligation, puncture-induced sepsis and endotoxin-induced acute lung injury in mice [101, 570].

The inflammatory microenvironment of the patient is not the only variable that must be taken into account. The dose and the delivery route of MSC, and the intrinsic diversity of their populations, are factors which are still under investigation. It is important to ascertain whether MSC obtained from different donors, regardless of age or health status, are functionally similar. It has been demonstrated that when the cytokines and growth factors released by MSC are compared, the proportion of bioactive factors is relatively consistent amongst donors [571]. Hence, the bioactive agents secreted may

162 represent the functional status of MSC. Whether MSC sourced from patients retain normal features and functions remains controversial. In a phase I clinical trial, BM- derived MSC from patients with Crohn’s disease shared identical characteristics with MSC from healthy donors, such as in vitro immunomodulatory capacities [572]. Expanded autologous MSC from patients with Crohn’s disease have been demonstrated to promote the differentiation of regulatory T cells in vivo. MSC isolated from aplastic anemia patients are capable of supporting the in vitro homeostasis and the in vivo repopulating function of CD34+ cells, with preserved immunosuppressive and anti- inflammatory properties [573]. However, controversies exist about the chondrogenic potential of MSC obtained from osteoarthritis patients [574] [575]. It is plausible that the discrepancies may be related to MSC being exposed to different patient microenvironment and/or anti-inflammatory drugs.

Another important issue is the route of administration of MSC. MSC have been administered both systemically and locally. The actual site at which MSC exert their potential immunoregulatory effects is still under investigation; it could be in the target organ or in the lymph nodes. As a result, it is difficult to justify, in terms of efficacy, administering MSC locally or systemically. In addition, although MSC express homing receptors, whether such receptors lead to systemically administered MSC arriving at the target organ remains controversial. The choice of the route for administration should be carefully evaluated, and is possibly dependent on the desired clinical effects.

Defining the correct dose to achieve optimal efficacy is another critical issue. In a clinical trial treating GvHD patients with different doses of MSC, there were no differences in efficacy or safety between doses [576]. Additionally, in the brain injury murine model, increasing the dose of MSC administered did not produce a demonstrably better therapeutic effect [577]. Cumulatively, these findings suggest that the minimum and maximum effective doses of MSC could be determined based on the disease model.

In HSCT, MSC have been shown to be beneficial for several indications. They can be

163 used to enhance haematopoietic engraftment, accelerate lymphocyte recovery, reduce the risk of graft failure, prevent and treat GvHD, as well as to repair tissue damage. Although the underlying mechanisms responsible for these effects remain to be elucidated, consideration of the role of MSC in facilitating the HSC niche may be a step towards identifying the possible mechanisms.

Environmental stimuli, such as inflammation, transplantations, bleeding, infections and chemotherapeutic agents stress the dormant HSC and increase their proliferation and differentiation. Meanwhile, quiescence represents a mechanism for minimising the accumulation of cellular damage induced by environmental stress as well as for preserving the life-long self-renewal capacity of HSC. Thus, it is plausible that the immunosuppressive effect of MSC, by inducing non-specific cell-cycle arrest, could regulate the response of HSC against stress according to specific inflammatory cues. It has also become evident that MSC regulate HSC mobilisation and egress, with bioenergetics and metabolism considered to play important roles in determining the fate of HSC. In the hypoxic niche, where MSC normally reside, resting quiescent HSC preferentially employ anaerobic glycolysis for energy production, with this metabolic route critical to maintaining a quiescent state [578]. Other metabolic pathways, involving lipid and nucleotide metabolism, have been identified as being restrictive to HSC entry into the cell cycle [504].

As metabolic pathways are known to be important in the maintenance of quiescent HSC populations and that EAA metabolism is a key mediator of MSC in arresting immune cells in the G0 phase, it is plausible that the coordination and regulation of EAA metabolism by MSC also impacts on HSPC maintenance. As WT MSC, but not Nos2- /- MSC, are capable of retaining more non-adherent HSPC in the G0 phase, my studies suggested that NOS2 may be an important molecule in the HSC niche. Interestingly, it has been shown that histamine prevents cell cycle induction via H4R activation on HSPC; this further supports the notion that metabolic pathways involved in controlling immune cell proliferation might overlap with those regulating HSC quiescence [579]. Although the role of EAA depletion in regulating HSC metabolism remains to be elucidated in detail, controlling HSC metabolism via MSC represents an attractive tool for the maintenance of the HSC niche, facilitating their engraftment, as well as sensitising malignant HSC to cytotoxic agents. 164

Striking parallels have been acknowledged between the normal and malignant stem cell niche. The characteristic of self-renewal of stem cells is shared with tumour cells, which reside in a highly inflammatory milieu. Due to the vast amount of chemokines and cytokines released by tumour cells, the tumour has been referred to as, ‘a wound that never heals’[580]. Tumour cells hijack normal tissue niches for their own survival by reconfiguring the microenvironment of the host. Most solid malignancies trigger an intrinsic inflammatory response that builds up a pro-tumourigenic microenvironment [581]. In addition to cell-autonomous proliferation, cancer cells are embedded in a framework where macrophages and fibroblasts foster their growth and promote immune evasion by inducing an immunosuppressive environment. In haematopoietic malignancies, it is well documented that leukaemia cells exploit the capacity of the niche to maintain the quiescence of malignant HSC and evade the anti-proliferative and pro-apoptotic effects of chemotherapeutic agents, accounting for the relapse of disease.

In the present study the functions of MSC across the fields of immunomodulation, niche-supporting and anti-apoptotic effects have been explored. It is anticipated that the findings will benefit the development of therapeutic treatments aimed at controlling inflammation, re-constructing the niche, or destroying cancer stem cells.

165

7 REFERENCES

1. Tulina, N. and E. Matunis, Control of stem cell self-renewal in Drosophila spermatogenesis by JAK-STAT signaling. Science, 2001. 294(5551): p. 2546-9. 2. Song, X. and T. Xie, DE-cadherin-mediated cell adhesion is essential for maintaining somatic stem cells in the Drosophila ovary. Proc Natl Acad Sci U S A, 2002. 99(23): p. 14813-8. 3. Winkler, I.G., et al., Vascular niche E-selectin regulates hematopoietic stem cell dormancy, self renewal and chemoresistance. Nat Med, 2012. 18(11): p. 1651-7. 4. Engler, A.J., et al., Matrix elasticity directs stem cell lineage specification. Cell, 2006. 126(4): p. 677-89. 5. Brizzi, M.F., G. Tarone, and P. Defilippi, Extracellular matrix, integrins, and growth factors as tailors of the stem cell niche. Curr Opin Cell Biol, 2012. 24(5): p. 645-51. 6. Yamashita, Y.M., D.L. Jones, and M.T. Fuller, Orientation of asymmetric stem cell division by the APC tumor suppressor and centrosome. Science, 2003. 301(5639): p. 1547-50. 7. Jones, D.L. and A.J. Wagers, No place like home: anatomy and function of the stem cell niche. Nat Rev Mol Cell Biol, 2008. 9(1): p. 11-21. 8. Dominici, M., et al., Minimal criteria for defining multipotent mesenchymal stromal cells. The International Society for Cellular Therapy position statement. Cytotherapy, 2006. 8(4): p. 315-7. 9. Luria, E.A., A.F. Panasyuk, and A.Y. Friedenstein, Fibroblast colony formation from monolayer cultures of blood cells. Transfusion, 1971. 11(6): p. 345-9. 10. Reyes, M., et al., Origin of endothelial progenitors in human postnatal bone marrow. J Clin Invest, 2002. 109(3): p. 337-46. 11. Cho, K.J., et al., Neurons derived from human mesenchymal stem cells show synaptic transmission and can be induced to produce the neurotransmitter substance P by interleukin-1 alpha. Stem Cells, 2005. 23(3): p. 383-91. 12. Greco, S.J., et al., An interdisciplinary approach and characterization of neuronal cells transdifferentiated from human mesenchymal stem cells. Stem Cells Dev, 2007. 16(5): p. 811-26. 13. Sato, Y., et al., Human mesenchymal stem cells xenografted directly to rat liver are differentiated into human hepatocytes without fusion. Blood, 2005. 106(2): p. 756-63. 14. Choi, K.S., et al., In vitro trans-differentiation of rat mesenchymal cells into insulin-producing cells by rat pancreatic extract. Biochem Biophys Res Commun, 2005. 330(4): p. 1299-305. 15. da Silva Meirelles, L., P.C. Chagastelles, and N.B. Nardi, Mesenchymal stem cells reside in virtually all post-natal organs and tissues. J Cell Sci, 2006. 119(Pt 11): p. 2204-13. 16. Zuk, P.A., et al., Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell, 2002. 13(12): p. 4279-95.

166

17. Toma, J.G., et al., Isolation of multipotent adult stem cells from the dermis of mammalian skin. Nat Cell Biol, 2001. 3(9): p. 778-84. 18. In 't Anker, P.S., et al., Isolation of mesenchymal stem cells of fetal or maternal origin from human placenta. Stem Cells, 2004. 22(7): p. 1338-45. 19. Hoogduijn, M.J., et al., Human heart, spleen, and perirenal fat-derived mesenchymal stem cells have immunomodulatory capacities. Stem Cells Dev, 2007. 16(4): p. 597-604. 20. Horwitz, E.M., et al., Clarification of the nomenclature for MSC: The International Society for Cellular Therapy position statement. Cytotherapy, 2005. 7(5): p. 393-5. 21. Muraglia, A., R. Cancedda, and R. Quarto, Clonal mesenchymal progenitors from human bone marrow differentiate in vitro according to a hierarchical model. J Cell Sci, 2000. 113 ( Pt 7): p. 1161-6. 22. Colter, D.C., I. Sekiya, and D.J. Prockop, Identification of a subpopulation of rapidly self-renewing and multipotential adult stem cells in colonies of human marrow stromal cells. Proc Natl Acad Sci U S A, 2001. 98(14): p. 7841-5. 23. Digirolamo, C.M., et al., Propagation and senescence of human marrow stromal cells in culture: a simple colony-forming assay identifies samples with the greatest potential to propagate and differentiate. Br J Haematol, 1999. 107(2): p. 275-81. 24. Kuznetsov, S.A., et al., Single-colony derived strains of human marrow stromal fibroblasts form bone after transplantation in vivo. J Bone Miner Res, 1997. 12(9): p. 1335-47. 25. Friedenstein, A.J., et al., Stromal cells responsible for transferring the microenvironment of the hemopoietic tissues. Cloning in vitro and retransplantation in vivo. Transplantation, 1974. 17(4): p. 331-40. 26. Gronthos, S., et al., Postnatal human dental pulp stem cells (DPSCs) in vitro and in vivo. Proc Natl Acad Sci U S A, 2000. 97(25): p. 13625-30. 27. Fukuchi, Y., et al., Human placenta-derived cells have mesenchymal stem/progenitor cell potential. Stem Cells, 2004. 22(5): p. 649-58. 28. Martins, A.A., et al., Quantification and immunophenotypic characterization of bone marrow and umbilical cord blood mesenchymal stem cells by multicolor flow cytometry. Transplant Proc, 2009. 41(3): p. 943-6. 29. Klingemann, H., D. Matzilevich, and J. Marchand, Mesenchymal Stem Cells - Sources and Clinical Applications. Transfus Med Hemother, 2008. 35(4): p. 272-277. 30. da Silva Meirelles, L., A.I. Caplan, and N.B. Nardi, In search of the in vivo identity of mesenchymal stem cells. Stem Cells, 2008. 26(9): p. 2287-99. 31. Jones, S., et al., The antiproliferative effect of mesenchymal stem cells is a fundamental property shared by all stromal cells. J Immunol, 2007. 179(5): p. 2824-31. 32. Takashima, Y., et al., Neuroepithelial cells supply an initial transient wave of MSC differentiation. Cell, 2007. 129(7): p. 1377-88. 33. Sacchetti, B., et al., Self-renewing osteoprogenitors in bone marrow sinusoids can organize a hematopoietic microenvironment. Cell, 2007. 131(2): p. 324-36. 34. Pan, G. and J.A. Thomson, Nanog and transcriptional networks in embryonic stem cell pluripotency. Cell Res, 2007. 17(1): p. 42-9.

167

35. Krampera, M., et al., Mesenchymal stem cells: from biology to clinical use. Blood Transfus, 2007. 5(3): p. 120-9. 36. Serafini, M., et al., Hematopoietic reconstitution by multipotent adult progenitor cells: precursors to long-term hematopoietic stem cells. J Exp Med, 2007. 204(1): p. 129-39. 37. Kucia, M., et al., A population of very small embryonic-like (VSEL) CXCR4(+)SSEA-1(+)Oct-4+ stem cells identified in adult bone marrow. Leukemia, 2006. 20(5): p. 857-69. 38. Vodyanik, M.A., et al., A mesoderm-derived precursor for mesenchymal stem and endothelial cells. Cell Stem Cell, 2010. 7(6): p. 718-29. 39. Crisan, M., et al., A perivascular origin for mesenchymal stem cells in multiple human organs. Cell Stem Cell, 2008. 3(3): p. 301-13. 40. Bexell, D., et al., Bone marrow multipotent mesenchymal stroma cells act as pericyte-like migratory vehicles in experimental gliomas. Mol Ther, 2009. 17(1): p. 183-90. 41. Pittenger, M.F., et al., Multilineage potential of adult human mesenchymal stem cells. Science, 1999. 284(5411): p. 143-7. 42. Simmons, P.J. and B. Torok-Storb, Identification of stromal cell precursors in human bone marrow by a novel monoclonal antibody, STRO-1. Blood, 1991. 78(1): p. 55-62. 43. Gronthos, S., et al., Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow. J Cell Sci, 2003. 116(Pt 9): p. 1827-35. 44. Morikawa, S., et al., Prospective identification, isolation, and systemic transplantation of multipotent mesenchymal stem cells in murine bone marrow. J Exp Med, 2009. 206(11): p. 2483-96. 45. Honczarenko, M., et al., Human bone marrow stromal cells express a distinct set of biologically functional chemokine receptors. Stem Cells, 2006. 24(4): p. 1030-41. 46. Leo, A.J. and D.A. Grande, Mesenchymal stem cells in tissue engineering. Cells Tissues Organs, 2006. 183(3): p. 112-22. 47. Ringe, J., et al., Towards in situ tissue repair: human mesenchymal stem cells express chemokine receptors CXCR1, CXCR2 and CCR2, and migrate upon stimulation with CXCL8 but not CCL2. J Cell Biochem, 2007. 101(1): p. 135- 46. 48. Wynn, R.F., et al., A small proportion of mesenchymal stem cells strongly expresses functionally active CXCR4 receptor capable of promoting migration to bone marrow. Blood, 2004. 104(9): p. 2643-5. 49. Sordi, V., et al., Bone marrow mesenchymal stem cells express a restricted set of functionally active chemokine receptors capable of promoting migration to pancreatic islets. Blood, 2005. 106(2): p. 419-27. 50. Ji, J.F., et al., Interactions of chemokines and chemokine receptors mediate the migration of mesenchymal stem cells to the impaired site in the brain after hypoglossal nerve injury. Stem Cells, 2004. 22(3): p. 415-27. 51. Chamberlain, G., et al., Murine mesenchymal stem cells exhibit a restricted repertoire of functional chemokine receptors: comparison with human. PLoS One, 2008. 3(8): p. e2934. 52. Janeway, C.A., Jr. and R. Medzhitov, Innate immune recognition. Annu Rev Immunol, 2002. 20: p. 197-216.

168

53. Akira, S., S. Uematsu, and O. Takeuchi, Pathogen recognition and innate immunity. Cell, 2006. 124(4): p. 783-801. 54. Akira, S., K. Takeda, and T. Kaisho, Toll-like receptors: critical proteins linking innate and acquired immunity. Nat Immunol, 2001. 2(8): p. 675-80. 55. Applequist, S.E., R.P. Wallin, and H.G. Ljunggren, Variable expression of Toll-like receptor in murine innate and adaptive immune cell lines. Int Immunol, 2002. 14(9): p. 1065-74. 56. O'Mahony, D.S., et al., Differential constitutive and cytokine-modulated expression of human Toll-like receptors in primary neutrophils, monocytes, and macrophages. Int J Med Sci, 2008. 5(1): p. 1-8. 57. Medzhitov, R., et al., MyD88 is an adaptor protein in the hToll/IL-1 receptor family signaling pathways. Mol Cell, 1998. 2(2): p. 253-8. 58. Akira, S. and K. Takeda, Toll-like receptor signalling. Nat Rev Immunol, 2004. 4(7): p. 499-511. 59. Dempsey, P.W., S.A. Vaidya, and G. Cheng, The art of war: Innate and adaptive immune responses. Cell Mol Life Sci, 2003. 60(12): p. 2604-21. 60. Banchereau, J. and R.M. Steinman, Dendritic cells and the control of immunity. Nature, 1998. 392(6673): p. 245-52. 61. Sozzani, S., et al., Differential regulation of chemokine receptors during dendritic cell maturation: a model for their trafficking properties. J Immunol, 1998. 161(3): p. 1083-6. 62. Moser, M. and K.M. Murphy, Dendritic cell regulation of TH1-TH2 development. Nat Immunol, 2000. 1(3): p. 199-205. 63. Maldonado-Lopez, R. and M. Moser, Dendritic cell subsets and the regulation of Th1/Th2 responses. Semin Immunol, 2001. 13(5): p. 275-82. 64. Romagnani, S., T-cell subsets (Th1 versus Th2). Ann Allergy Asthma Immunol, 2000. 85(1): p. 9-18; quiz 18, 21. 65. Chen, C.J., et al., Identification of a key pathway required for the sterile inflammatory response triggered by dying cells. Nat Med, 2007. 13(7): p. 851-6. 66. Savill, J., et al., A blast from the past: clearance of apoptotic cells regulates immune responses. Nat Rev Immunol, 2002. 2(12): p. 965-75. 67. Marigo, I. and F. Dazzi, The immunomodulatory properties of mesenchymal stem cells. Semin Immunopathol, 2011. 33(6): p. 593-602. 68. Nauta, A.J. and W.E. Fibbe, Immunomodulatory properties of mesenchymal stromal cells. Blood, 2007. 110(10): p. 3499-506. 69. Abdi, R., et al., Immunomodulation by mesenchymal stem cells: a potential therapeutic strategy for type 1 diabetes. Diabetes, 2008. 57(7): p. 1759-67. 70. Bassi, E.J., C.A. Aita, and N.O. Camara, Immune regulatory properties of multipotent mesenchymal stromal cells: Where do we stand? World J Stem Cells, 2011. 3(1): p. 1-8. 71. Malhotra, D., et al., Transcriptional profiling of stroma from inflamed and resting lymph nodes defines immunological hallmarks. Nat Immunol, 2012. 13(5): p. 499-510. 72. Shi, Y., et al., How mesenchymal stem cells interact with tissue immune responses. Trends Immunol, 2012. 33(3): p. 136-43. 73. Krampera, M., Mesenchymal stromal cell 'licensing': a multistep process. Leukemia, 2011. 25(9): p. 1408-14.

169

74. Prockop, D.J. and J.Y. Oh, Mesenchymal stem/stromal cells (MSCs): role as guardians of inflammation. Mol Ther, 2012. 20(1): p. 14-20. 75. Krampera, M., et al., Role for interferon-gamma in the immunomodulatory activity of human bone marrow mesenchymal stem cells. Stem Cells, 2006. 24(2): p. 386-98. 76. Mougiakakos, D., et al., The impact of inflammatory licensing on heme oxygenase-1-mediated induction of regulatory T cells by human mesenchymal stem cells. Blood, 2011. 117(18): p. 4826-35. 77. Keating, A., How do mesenchymal stromal cells suppress T cells? Cell Stem Cell, 2008. 2(2): p. 106-8. 78. Dazzi, F. and F.M. Marelli-Berg, Mesenchymal stem cells for graft-versus- host disease: close encounters with T cells. Eur J Immunol, 2008. 38(6): p. 1479-82. 79. Ren, G., et al., Mesenchymal stem cell-mediated immunosuppression occurs via concerted action of chemokines and nitric oxide. Cell Stem Cell, 2008. 2(2): p. 141-50. 80. Hemeda, H., et al., Interferon-gamma and tumor necrosis factor-alpha differentially affect cytokine expression and migration properties of mesenchymal stem cells. Stem Cells Dev, 2010. 19(5): p. 693-706. 81. Liotta, F., et al., Toll-like receptors 3 and 4 are expressed by human bone marrow-derived mesenchymal stem cells and can inhibit their T-cell modulatory activity by impairing Notch signaling. Stem Cells, 2008. 26(1): p. 279-89. 82. DelaRosa, O. and E. Lombardo, Modulation of adult mesenchymal stem cells activity by toll-like receptors: implications on therapeutic potential. Mediators Inflamm, 2010. 2010: p. 865601. 83. Tomchuck, S.L., et al., Toll-like receptors on human mesenchymal stem cells drive their migration and immunomodulating responses. Stem Cells, 2008. 26(1): p. 99-107. 84. Romieu-Mourez, R., et al., Cytokine modulation of TLR expression and activation in mesenchymal stromal cells leads to a proinflammatory phenotype. J Immunol, 2009. 182(12): p. 7963-73. 85. Kim, H.S., et al., Implication of NOD1 and NOD2 for the differentiation of multipotent mesenchymal stem cells derived from human umbilical cord blood. PLoS One, 2010. 5(10): p. e15369. 86. Kume, S., et al., Advanced glycation end-products attenuate human mesenchymal stem cells and prevent cognate differentiation into adipose tissue, cartilage, and bone. J Bone Miner Res, 2005. 20(9): p. 1647-58. 87. Opitz, C.A., et al., Toll-like receptor engagement enhances the immunosuppressive properties of human bone marrow-derived mesenchymal stem cells by inducing indoleamine-2,3-dioxygenase-1 via interferon-beta and protein kinase R. Stem Cells, 2009. 27(4): p. 909-19. 88. Waterman, R.S., et al., A new mesenchymal stem cell (MSC) paradigm: polarization into a pro-inflammatory MSC1 or an Immunosuppressive MSC2 phenotype. PLoS One, 2010. 5(4): p. e10088. 89. de Oliveira, G.L., et al., Bone marrow mesenchymal stromal cells isolated from multiple sclerosis patients have distinct gene expression profile and decreased suppressive function compared with healthy counterparts. Cell Transplant, 2013.

170

90. Ren, G., et al., Tumor resident mesenchymal stromal cells endow naive stromal cells with tumor-promoting properties. Oncogene, 2013. 91. Soehnlein, O. and L. Lindbom, Phagocyte partnership during the onset and resolution of inflammation. Nat Rev Immunol, 2010. 10(6): p. 427-39. 92. De Filippo, K., et al., Mast cell and macrophage chemokines CXCL1/CXCL2 control the early stage of neutrophil recruitment during tissue inflammation. Blood, 2013. 121(24): p. 4930-7. 93. Marks, S.M., Z. Taylor, and B.I. Miller, Tuberculosis prevention versus hospitalization: taxpayers save with prevention. J Health Care Poor Underserved, 2002. 13(3): p. 392-401. 94. Serbina, N.V., et al., Monocyte-mediated defense against microbial pathogens. Annu Rev Immunol, 2008. 26: p. 421-52. 95. Brandau, S., et al., Tissue-resident mesenchymal stem cells attract peripheral blood neutrophils and enhance their inflammatory activity in response to microbial challenge. J Leukoc Biol, 2010. 88(5): p. 1005-15. 96. Yang, D., et al., Human neutrophil defensins selectively chemoattract naive T and immature dendritic cells. J Leukoc Biol, 2000. 68(1): p. 9-14. 97. Chertov, O., et al., Identification of human neutrophil-derived cathepsin G and azurocidin/CAP37 as chemoattractants for mononuclear cells and neutrophils. J Exp Med, 1997. 186(5): p. 739-47. 98. Shi, C., et al., Bone marrow mesenchymal stem and progenitor cells induce monocyte emigration in response to circulating toll-like receptor ligands. Immunity, 2011. 34(4): p. 590-601. 99. Chen, L., et al., Paracrine factors of mesenchymal stem cells recruit macrophages and endothelial lineage cells and enhance wound healing. PLoS One, 2008. 3(4): p. e1886. 100. Liu, L., et al., Human umbilical cord mesenchymal stem cells transplantation promotes cutaneous wound healing of severe burned rats. PLoS One, 2014. 9(2): p. e88348. 101. Nemeth, K., et al., Bone marrow stromal cells attenuate sepsis via prostaglandin E(2)-dependent reprogramming of host macrophages to increase their interleukin-10 production. Nat Med, 2009. 15(1): p. 42-9. 102. Dyer, D.P., et al., TSG-6 inhibits neutrophil migration via direct interaction with the chemokine CXCL8. J Immunol, 2014. 192(5): p. 2177-85. 103. Roddy, G.W., et al., Action at a distance: systemically administered adult stem/progenitor cells (MSCs) reduce inflammatory damage to the cornea without engraftment and primarily by secretion of TNF-alpha stimulated gene/protein 6. Stem Cells, 2011. 29(10): p. 1572-9. 104. Choi, H., et al., Anti-inflammatory protein TSG-6 secreted by activated MSCs attenuates zymosan-induced mouse peritonitis by decreasing TLR2/NF- kappaB signaling in resident macrophages. Blood, 2011. 118(2): p. 330-8. 105. Bernardo, M.E. and W.E. Fibbe, Mesenchymal stromal cells: sensors and switchers of inflammation. Cell Stem Cell, 2013. 13(4): p. 392-402. 106. Bai, L., et al., Human bone marrow-derived mesenchymal stem cells induce Th2-polarized immune response and promote endogenous repair in animal models of multiple sclerosis. Glia, 2009. 57(11): p. 1192-203. 107. Boumaza, I., et al., Autologous bone marrow-derived rat mesenchymal stem cells promote PDX-1 and insulin expression in the islets, alter T cell cytokine

171

pattern and preserve regulatory T cells in the periphery and induce sustained normoglycemia. J Autoimmun, 2009. 32(1): p. 33-42. 108. Madec, A.M., et al., Mesenchymal stem cells protect NOD mice from diabetes by inducing regulatory T cells. Diabetologia, 2009. 52(7): p. 1391-9. 109. Ghannam, S., et al., Mesenchymal stem cells inhibit human Th17 cell differentiation and function and induce a T regulatory cell phenotype. J Immunol, 2010. 185(1): p. 302-12. 110. Wang, J., et al., Interleukin-27 suppresses experimental autoimmune encephalomyelitis during bone marrow stromal cell treatment. J Autoimmun, 2008. 30(4): p. 222-9. 111. Rafei, M., et al., Mesenchymal stromal cells ameliorate experimental autoimmune encephalomyelitis by inhibiting CD4 Th17 T cells in a CC chemokine ligand 2-dependent manner. J Immunol, 2009. 182(10): p. 5994-6002. 112. Zappia, E., et al., Mesenchymal stem cells ameliorate experimental autoimmune encephalomyelitis inducing T-cell anergy. Blood, 2005. 106(5): p. 1755-61. 113. Zhao, W., et al., TGF-beta expression by allogeneic bone marrow stromal cells ameliorates diabetes in NOD mice through modulating the distribution of CD4+ T cell subsets. Cell Immunol, 2008. 253(1-2): p. 23-30. 114. Kong, Q.F., et al., BM stromal cells ameliorate experimental autoimmune myasthenia gravis by altering the balance of Th cells through the secretion of IDO. Eur J Immunol, 2009. 39(3): p. 800-9. 115. Duffy, M.M., et al., Mesenchymal stem cells and a vitamin D receptor agonist additively suppress T-helper 17 cells and related inflammatory response in the kidney. Am J Physiol Renal Physiol, 2014: p. ajprenal 00024 2014. 116. Nemeth, K., et al., Bone marrow stromal cells use TGF-beta to suppress allergic responses in a mouse model of ragweed-induced asthma. Proc Natl Acad Sci U S A, 2010. 107(12): p. 5652-7. 117. Curotto de Lafaille, M.A. and J.J. Lafaille, Natural and adaptive foxp3+ regulatory T cells: more of the same or a division of labor? Immunity, 2009. 30(5): p. 626-35. 118. Bluestone, J.A. and A.K. Abbas, Natural versus adaptive regulatory T cells. Nat Rev Immunol, 2003. 3(3): p. 253-7. 119. Maccario, R., et al., Interaction of human mesenchymal stem cells with cells involved in alloantigen-specific immune response favors the differentiation of CD4+ T-cell subsets expressing a regulatory/suppressive phenotype. Haematologica, 2005. 90(4): p. 516-25. 120. Aggarwal, S. and M.F. Pittenger, Human mesenchymal stem cells modulate allogeneic immune cell responses. Blood, 2005. 105(4): p. 1815-22. 121. Di Ianni, M., et al., Mesenchymal cells recruit and regulate T regulatory cells. Exp Hematol, 2008. 36(3): p. 309-18. 122. English, K., et al., Cell contact, prostaglandin E(2) and transforming growth factor beta 1 play non-redundant roles in human mesenchymal stem cell induction of CD4+CD25(High) forkhead box P3+ regulatory T cells. Clin Exp Immunol, 2009. 156(1): p. 149-60. 123. Prevosto, C., et al., Generation of CD4+ or CD8+ regulatory T cells upon mesenchymal stem cell-lymphocyte interaction. Haematologica, 2007. 92(7): p. 881-8.

172

124. Samon, J.B., et al., Notch1 and TGFbeta1 cooperatively regulate Foxp3 expression and the maintenance of peripheral regulatory T cells. Blood, 2008. 112(5): p. 1813-21. 125. Del Papa, B., et al., Notch1 modulates mesenchymal stem cells mediated regulatory T-cell induction. Eur J Immunol, 2013. 43(1): p. 182-7. 126. Selmani, Z., et al., Human leukocyte antigen-G5 secretion by human mesenchymal stem cells is required to suppress T lymphocyte and natural killer function and to induce CD4+CD25highFOXP3+ regulatory T cells. Stem Cells, 2008. 26(1): p. 212-22. 127. Gonzalez, M.A., et al., Adipose-derived mesenchymal stem cells alleviate experimental colitis by inhibiting inflammatory and autoimmune responses. Gastroenterology, 2009. 136(3): p. 978-89. 128. Kavanagh, H. and B.P. Mahon, Allogeneic mesenchymal stem cells prevent allergic airway inflammation by inducing murine regulatory T cells. Allergy, 2011. 66(4): p. 523-31. 129. Ge, W., et al., Regulatory T-cell generation and kidney allograft tolerance induced by mesenchymal stem cells associated with indoleamine 2,3- dioxygenase expression. Transplantation, 2010. 90(12): p. 1312-20. 130. Wang, Y., et al., Bone marrow-derived mesenchymal stem cells inhibit acute rejection of rat liver allografts in association with regulatory T-cell expansion. Transplant Proc, 2009. 41(10): p. 4352-6. 131. Casiraghi, F., et al., Pretransplant infusion of mesenchymal stem cells prolongs the survival of a semiallogeneic heart transplant through the generation of regulatory T cells. J Immunol, 2008. 181(6): p. 3933-46. 132. Ge, W., et al., Infusion of mesenchymal stem cells and rapamycin synergize to attenuate alloimmune responses and promote cardiac allograft tolerance. Am J Transplant, 2009. 9(8): p. 1760-72. 133. Clark, G.J., et al., The role of dendritic cells in the innate immune system. Microbes Infect, 2000. 2(3): p. 257-72. 134. Wallet, M.A., P. Sen, and R. Tisch, Immunoregulation of dendritic cells. Clin Med Res, 2005. 3(3): p. 166-75. 135. Mahnke, K., et al., Immature, but not inactive: the tolerogenic function of immature dendritic cells. Immunol Cell Biol, 2002. 80(5): p. 477-83. 136. Ramasamy, R., et al., Mesenchymal stem cells inhibit dendritic cell differentiation and function by preventing entry into the cell cycle. Transplantation, 2007. 83(1): p. 71-6. 137. Zhang, W., et al., Effects of mesenchymal stem cells on differentiation, maturation, and function of human monocyte-derived dendritic cells. Stem Cells Dev, 2004. 13(3): p. 263-71. 138. Spaggiari, G.M., et al., MSCs inhibit monocyte-derived DC maturation and function by selectively interfering with the generation of immature DCs: central role of MSC-derived prostaglandin E2. Blood, 2009. 113(26): p. 6576-83. 139. Djouad, F., et al., Mesenchymal stem cells inhibit the differentiation of dendritic cells through an interleukin-6-dependent mechanism. Stem Cells, 2007. 25(8): p. 2025-32. 140. English, K., F.P. Barry, and B.P. Mahon, Murine mesenchymal stem cells suppress dendritic cell migration, maturation and antigen presentation. Immunol Lett, 2008. 115(1): p. 50-8.

173

141. Banchereau, J., et al., Immunobiology of dendritic cells. Annu Rev Immunol, 2000. 18: p. 767-811. 142. Mellman, I. and R.M. Steinman, Dendritic cells: specialized and regulated antigen processing machines. Cell, 2001. 106(3): p. 255-8. 143. Wang, Q., et al., Murine bone marrow mesenchymal stem cells cause mature dendritic cells to promote T-cell tolerance. Scand J Immunol, 2008. 68(6): p. 607-15. 144. Li, Y.P., et al., Human mesenchymal stem cells license adult CD34+ hemopoietic progenitor cells to differentiate into regulatory dendritic cells through activation of the Notch pathway. J Immunol, 2008. 180(3): p. 1598-608. 145. Popp, F.C., et al., Mesenchymal stem cells can induce long-term acceptance of solid organ allografts in synergy with low-dose mycophenolate. Transpl Immunol, 2008. 20(1-2): p. 55-60. 146. Choi, Y.S., J.A. Jeong, and D.S. Lim, Mesenchymal stem cell-mediated immature dendritic cells induce regulatory T cell-based immunosuppressive effect. Immunol Invest, 2012. 41(2): p. 214-29. 147. Chiesa, S., et al., Mesenchymal stem cells impair in vivo T-cell priming by dendritic cells. Proc Natl Acad Sci U S A, 2011. 108(42): p. 17384-9. 148. Wang, Z., et al., The different immunoregulatory functions on dendritic cells between mesenchymal stem cells derived from bone marrow of patients with low-risk or high-risk myelodysplastic syndromes. PLoS One, 2013. 8(3): p. e57470. 149. Zhao, Z.G., et al., The characteristics and immunoregulatory functions of regulatory dendritic cells induced by mesenchymal stem cells derived from bone marrow of patient with chronic myeloid leukaemia. Eur J Cancer, 2012. 48(12): p. 1884-95. 150. Gordon, S., Alternative activation of macrophages. Nat Rev Immunol, 2003. 3(1): p. 23-35. 151. Mills, C.D., et al., M-1/M-2 macrophages and the Th1/Th2 paradigm. J Immunol, 2000. 164(12): p. 6166-73. 152. Mantovani, A., et al., The chemokine system in diverse forms of macrophage activation and polarization. Trends Immunol, 2004. 25(12): p. 677-86. 153. Francois, M., et al., Human MSC suppression correlates with cytokine induction of indoleamine 2,3-dioxygenase and bystander M2 macrophage differentiation. Mol Ther, 2012. 20(1): p. 187-95. 154. Zhang, Q.Z., et al., Human gingiva-derived mesenchymal stem cells elicit polarization of m2 macrophages and enhance cutaneous wound healing. Stem Cells, 2010. 28(10): p. 1856-68. 155. Cho, D.I., et al., Mesenchymal stem cells reciprocally regulate the M1/M2 balance in mouse bone marrow-derived macrophages. Exp Mol Med, 2014. 46: p. e70. 156. Nakajima, H., et al., Transplantation of mesenchymal stem cells promotes an alternative pathway of macrophage activation and functional recovery after spinal cord injury. J Neurotrauma, 2012. 29(8): p. 1614-25. 157. Melief, S.M., et al., Multipotent stromal cells induce human regulatory T cells through a novel pathway involving skewing of monocytes toward anti- inflammatory macrophages. Stem Cells, 2013. 31(9): p. 1980-91.

174

158. Krampera, M., et al., Bone marrow mesenchymal stem cells inhibit the response of naive and memory antigen-specific T cells to their cognate peptide. Blood, 2003. 101(9): p. 3722-9. 159. Djouad, F., et al., Immunosuppressive effect of mesenchymal stem cells favors tumor growth in allogeneic animals. Blood, 2003. 102(10): p. 3837- 44. 160. Bartholomew, A., et al., Mesenchymal stem cells suppress lymphocyte proliferation in vitro and prolong skin graft survival in vivo. Exp Hematol, 2002. 30(1): p. 42-8. 161. Potian, J.A., et al., Veto-like activity of mesenchymal stem cells: functional discrimination between cellular responses to alloantigens and recall antigens. J Immunol, 2003. 171(7): p. 3426-34. 162. Tse, W.T., et al., Suppression of allogeneic T-cell proliferation by human marrow stromal cells: implications in transplantation. Transplantation, 2003. 75(3): p. 389-97. 163. Di Nicola, M., et al., Human bone marrow stromal cells suppress T- lymphocyte proliferation induced by cellular or nonspecific mitogenic stimuli. Blood, 2002. 99(10): p. 3838-43. 164. Najar, M., et al., Mesenchymal stromal cells promote or suppress the proliferation of T lymphocytes from cord blood and peripheral blood: the importance of low cell ratio and role of interleukin-6. Cytotherapy, 2009. 11(5): p. 570-83. 165. Yang, S.H., et al., Soluble mediators from mesenchymal stem cells suppress T cell proliferation by inducing IL-10. Exp Mol Med, 2009. 41(5): p. 315-24. 166. Meisel, R., et al., Human bone marrow stromal cells inhibit allogeneic T-cell responses by indoleamine 2,3-dioxygenase-mediated tryptophan degradation. Blood, 2004. 103(12): p. 4619-21. 167. Sato, K., et al., Nitric oxide plays a critical role in suppression of T-cell proliferation by mesenchymal stem cells. Blood, 2007. 109(1): p. 228-34. 168. Lepelletier, Y., et al., Galectin-1 and semaphorin-3A are two soluble factors conferring T-cell immunosuppression to bone marrow mesenchymal stem cell. Stem Cells Dev, 2010. 19(7): p. 1075-9. 169. Gieseke, F., et al., Proinflammatory stimuli induce galectin-9 in human mesenchymal stromal cells to suppress T-cell proliferation. Eur J Immunol, 2013. 43(10): p. 2741-9. 170. Ren, G., et al., Inflammatory cytokine-induced intercellular adhesion molecule-1 and vascular cell adhesion molecule-1 in mesenchymal stem cells are critical for immunosuppression. J Immunol, 2010. 184(5): p. 2321-8. 171. Chinnadurai, R., et al., IDO-independent suppression of T cell effector function by IFN-gamma-licensed human mesenchymal stromal cells. J Immunol, 2014. 192(4): p. 1491-501. 172. Augello, A., et al., Bone marrow mesenchymal progenitor cells inhibit lymphocyte proliferation by activation of the programmed death 1 pathway. Eur J Immunol, 2005. 35(5): p. 1482-90. 173. Glennie, S., et al., Bone marrow mesenchymal stem cells induce division arrest anergy of activated T cells. Blood, 2005. 105(7): p. 2821-7. 174. Wells, A.D., et al., Signaling through CD28 and CTLA-4 controls two distinct forms of T cell anergy. J Clin Invest, 2001. 108(6): p. 895-903.

175

175. Klyushnenkova, E., et al., T cell responses to allogeneic human mesenchymal stem cells: immunogenicity, tolerance, and suppression. J Biomed Sci, 2005. 12(1): p. 47-57. 176. Kurosaka, D., T.W. LeBien, and J.A. Pribyl, Comparative studies of different stromal cell microenvironments in support of human B-cell development. Exp Hematol, 1999. 27(8): p. 1271-81. 177. Corcione, A., et al., Human mesenchymal stem cells modulate B-cell functions. Blood, 2006. 107(1): p. 367-72. 178. Deng, W., et al., Effects of allogeneic bone marrow-derived mesenchymal stem cells on T and B lymphocytes from BXSB mice. DNA Cell Biol, 2005. 24(7): p. 458-63. 179. Spaggiari, G.M., et al., Mesenchymal stem cell-natural killer cell interactions: evidence that activated NK cells are capable of killing MSCs, whereas MSCs can inhibit IL-2-induced NK-cell proliferation. Blood, 2006. 107(4): p. 1484-90. 180. Sotiropoulou, P.A., et al., Interactions between human mesenchymal stem cells and natural killer cells. Stem Cells, 2006. 24(1): p. 74-85. 181. Spaggiari, G.M., et al., Mesenchymal stem cells inhibit natural killer-cell proliferation, cytotoxicity, and cytokine production: role of indoleamine 2,3- dioxygenase and prostaglandin E2. Blood, 2008. 111(3): p. 1327-33. 182. Grohmann, U. and V. Bronte, Control of immune response by amino acid metabolism. Immunol Rev, 2010. 236: p. 243-64. 183. Cobbold, S.P., et al., Infectious tolerance via the consumption of essential amino acids and mTOR signaling. Proc Natl Acad Sci U S A, 2009. 106(29): p. 12055-60. 184. von Bubnoff, D., et al., FcepsilonRI induces the tryptophan degradation pathway involved in regulating T cell responses. J Immunol, 2002. 169(4): p. 1810-6. 185. Matlack, R., et al., Peritoneal macrophages suppress T-cell activation by amino acid catabolism. Immunology, 2006. 117(3): p. 386-95. 186. Rodriguez, P.C. and A.C. Ochoa, Arginine regulation by myeloid derived suppressor cells and tolerance in cancer: mechanisms and therapeutic perspectives. Immunol Rev, 2008. 222: p. 180-91. 187. Appleton, J., Arginine: Clinical potential of a semi-essential amino acid. Altern Med Rev, 2002. 7(6): p. 512-22. 188. Wu, G. and S.M. Morris, Jr., Arginine metabolism: nitric oxide and beyond. Biochem J, 1998. 336 ( Pt 1): p. 1-17. 189. Morris, S.M., Jr., Arginine metabolism: boundaries of our knowledge. J Nutr, 2007. 137(6 Suppl 2): p. 1602S-1609S. 190. Field, C.J., I. Johnson, and V.C. Pratt, Glutamine and arginine: immunonutrients for improved health. Med Sci Sports Exerc, 2000. 32(7 Suppl): p. S377-88. 191. Jenkinson, C.P., W.W. Grody, and S.D. Cederbaum, Comparative properties of arginases. Comp Biochem Physiol B Biochem Mol Biol, 1996. 114(1): p. 107-32. 192. Ash, D.E., Structure and function of arginases. J Nutr, 2004. 134(10 Suppl): p. 2760S-2764S; discussion 2765S-2767S.

176

193. El Kasmi, K.C., et al., Toll-like receptor-induced arginase 1 in macrophages thwarts effective immunity against intracellular pathogens. Nat Immunol, 2008. 9(12): p. 1399-406. 194. Corraliza, I.M., et al., Arginase induction by suppressors of nitric oxide synthesis (IL-4, IL-10 and PGE2) in murine bone-marrow-derived macrophages. Biochem Biophys Res Commun, 1995. 206(2): p. 667-73. 195. Munder, M., K. Eichmann, and M. Modolell, Alternative metabolic states in murine macrophages reflected by the nitric oxide synthase/arginase balance: competitive regulation by CD4+ T cells correlates with Th1/Th2 phenotype. J Immunol, 1998. 160(11): p. 5347-54. 196. Louis, C.A., et al., Regulation of arginase isoforms I and II by IL-4 in cultured murine peritoneal macrophages. Am J Physiol, 1999. 276(1 Pt 2): p. R237- 42. 197. Munder, M., et al., Th1/Th2-regulated expression of arginase isoforms in murine macrophages and dendritic cells. J Immunol, 1999. 163(7): p. 3771-7. 198. Munder, M., et al., Arginase I is constitutively expressed in human granulocytes and participates in fungicidal activity. Blood, 2005. 105(6): p. 2549-56. 199. Morris, S.M., Jr., D. Kepka-Lenhart, and L.C. Chen, Differential regulation of arginases and inducible nitric oxide synthase in murine macrophage cells. Am J Physiol, 1998. 275(5 Pt 1): p. E740-7. 200. Grandvaux, N., et al., Regulation of arginase II by interferon regulatory factor 3 and the involvement of polyamines in the antiviral response. FEBS J, 2005. 272(12): p. 3120-31. 201. Corraliza, I. and S. Moncada, Increased expression of arginase II in patients with different forms of arthritis. Implications of the regulation of nitric oxide. J Rheumatol, 2002. 29(11): p. 2261-5. 202. Barksdale, A.R., et al., Regulation of arginase expression by T-helper II cytokines and isoproterenol. Surgery, 2004. 135(5): p. 527-35. 203. Fujita, K., et al., Nitric oxide plays a crucial role in the development/progression of nonalcoholic steatohepatitis in the choline- deficient, l-amino acid-defined diet-fed rat model. Alcohol Clin Exp Res, 2010. 34 Suppl 1: p. S18-24. 204. Rodriguez, P.C., et al., L-arginine consumption by macrophages modulates the expression of CD3 zeta chain in T lymphocytes. J Immunol, 2003. 171(3): p. 1232-9. 205. Rodriguez, P.C., et al., Arginase I production in the tumor microenvironment by mature myeloid cells inhibits T-cell receptor expression and antigen-specific T-cell responses. Cancer Res, 2004. 64(16): p. 5839-49. 206. Zea, A.H., et al., Arginase-producing myeloid suppressor cells in renal cell carcinoma patients: a mechanism of tumor evasion. Cancer Res, 2005. 65(8): p. 3044-8. 207. Tate, D.J., Jr., et al., Effect of arginase II on L-arginine depletion and cell growth in murine cell lines of renal cell carcinoma. J Hematol Oncol, 2008. 1: p. 14. 208. Coqueret, O., Linking cyclins to transcriptional control. Gene, 2002. 299(1- 2): p. 35-55.

177

209. Nevins, J.R., E2F: a link between the Rb tumor suppressor protein and viral oncoproteins. Science, 1992. 258(5081): p. 424-9. 210. Rodriguez, P.C., D.G. Quiceno, and A.C. Ochoa, L-arginine availability regulates T-lymphocyte cell-cycle progression. Blood, 2007. 109(4): p. 1568-73. 211. Bronte, V., et al., IL-4-induced arginase 1 suppresses alloreactive T cells in tumor-bearing mice. J Immunol, 2003. 170(1): p. 270-8. 212. Bronte, V. and P. Zanovello, Regulation of immune responses by L-arginine metabolism. Nat Rev Immunol, 2005. 5(8): p. 641-54. 213. Molon, B., et al., Chemokine nitration prevents intratumoral infiltration of antigen-specific T cells. J Exp Med, 2011. 208(10): p. 1949-62. 214. Brito, C., et al., Peroxynitrite inhibits T lymphocyte activation and proliferation by promoting impairment of tyrosine phosphorylation and peroxynitrite-driven apoptotic death. J Immunol, 1999. 162(6): p. 3356-66. 215. Hernandez, C.P., et al., Pegylated arginase I: a potential therapeutic approach in T-ALL. Blood, 2010. 115(25): p. 5214-21. 216. Highfill, S.L., et al., Bone marrow myeloid-derived suppressor cells (MDSCs) inhibit graft-versus-host disease (GVHD) via an arginase-1-dependent mechanism that is up-regulated by interleukin-13. Blood, 2010. 116(25): p. 5738-47. 217. Alderton, W.K., C.E. Cooper, and R.G. Knowles, Nitric oxide synthases: structure, function and inhibition. Biochem J, 2001. 357(Pt 3): p. 593-615. 218. Yun, H.Y., V.L. Dawson, and T.M. Dawson, Nitric oxide in health and disease of the nervous system. Mol Psychiatry, 1997. 2(4): p. 300-10. 219. Toda, N., K. Ayajiki, and T. Okamura, Control of systemic and pulmonary blood pressure by nitric oxide formed through neuronal nitric oxide synthase. J Hypertens, 2009. 27(10): p. 1929-40. 220. Adler, H.S., et al., Neuronal nitric oxide synthase modulates maturation of human dendritic cells. J Immunol, 2010. 184(11): p. 6025-34. 221. Huang, P.L., et al., Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature, 1995. 377(6546): p. 239-42. 222. Murohara, T., et al., Nitric oxide synthase modulates angiogenesis in response to tissue ischemia. J Clin Invest, 1998. 101(11): p. 2567-78. 223. Rudic, R.D., et al., Direct evidence for the importance of endothelium- derived nitric oxide in vascular remodeling. J Clin Invest, 1998. 101(4): p. 731-6. 224. Lee, P.C., et al., Impaired wound healing and angiogenesis in eNOS-deficient mice. Am J Physiol, 1999. 277(4 Pt 2): p. H1600-8. 225. Lin, L.Y., et al., Angiotensin II-induced apoptosis in human endothelial cells is inhibited by adiponectin through restoration of the association between endothelial nitric oxide synthase and heat shock protein 90. FEBS Lett, 2004. 574(1-3): p. 106-10. 226. Kanki-Horimoto, S., et al., Implantation of mesenchymal stem cells overexpressing endothelial nitric oxide synthase improves right ventricular impairments caused by pulmonary hypertension. Circulation, 2006. 114(1 Suppl): p. I181-5. 227. Bucci, M., et al., In vivo delivery of the caveolin-1 scaffolding domain inhibits nitric oxide synthesis and reduces inflammation. Nat Med, 2000. 6(12): p. 1362-7.

178

228. Chan, E.D. and D.W. Riches, IFN-gamma + LPS induction of iNOS is modulated by ERK, JNK/SAPK, and p38(mapk) in a mouse macrophage cell line. Am J Physiol Cell Physiol, 2001. 280(3): p. C441-50. 229. Xie, Q.W., Y. Kashiwabara, and C. Nathan, Role of transcription factor NF- kappa B/Rel in induction of nitric oxide synthase. J Biol Chem, 1994. 269(7): p. 4705-8. 230. Pautz, A., et al., Regulation of the expression of inducible nitric oxide synthase. Nitric Oxide, 2010. 23(2): p. 75-93. 231. Lowenstein, C.J. and E. Padalko, iNOS (NOS2) at a glance. J Cell Sci, 2004. 117(Pt 14): p. 2865-7. 232. Moelants, E.A., et al., Regulation of TNF-alpha with a focus on rheumatoid arthritis. Immunol Cell Biol, 2013. 91(6): p. 393-401. 233. Kamijo, R., et al., Requirement for transcription factor IRF-1 in NO synthase induction in macrophages. Science, 1994. 263(5153): p. 1612-5. 234. Saura, M., et al., Interaction of interferon regulatory factor-1 and nuclear factor kappaB during activation of inducible nitric oxide synthase transcription. J Mol Biol, 1999. 289(3): p. 459-71. 235. Thomas, D.D., et al., The chemical biology of nitric oxide: implications in cellular signaling. Free Radic Biol Med, 2008. 45(1): p. 18-31. 236. Connelly, L., et al., Biphasic regulation of NF-kappa B activity underlies the pro- and anti-inflammatory actions of nitric oxide. J Immunol, 2001. 166(6): p. 3873-81. 237. Reynaert, N.L., et al., Nitric oxide represses inhibitory kappaB kinase through S-nitrosylation. Proc Natl Acad Sci U S A, 2004. 101(24): p. 8945- 50. 238. Matthews, J.R., et al., Inhibition of NF-kappaB DNA binding by nitric oxide. Nucleic Acids Res, 1996. 24(12): p. 2236-42. 239. Chung, H.T., et al., Interactive relations between nitric oxide (NO) and carbon monoxide (CO): heme oxygenase-1/CO pathway is a key modulator in NO-mediated antiapoptosis and anti-inflammation. Methods Enzymol, 2008. 441: p. 329-38. 240. Polte, T., et al., Heme oxygenase-1 is a cGMP-inducible endothelial protein and mediates the cytoprotective action of nitric oxide. Arterioscler Thromb Vasc Biol, 2000. 20(5): p. 1209-15. 241. Bingisser, R.M., et al., Macrophage-derived nitric oxide regulates T cell activation via reversible disruption of the Jak3/STAT5 signaling pathway. J Immunol, 1998. 160(12): p. 5729-34. 242. Cifone, M.G., S. Ulisse, and A. Santoni, Natural killer cells and nitric oxide. Int Immunopharmacol, 2001. 1(8): p. 1513-24. 243. Diefenbach, A., et al., Type 1 interferon (IFNalpha/beta) and type 2 nitric oxide synthase regulate the innate immune response to a protozoan parasite. Immunity, 1998. 8(1): p. 77-87. 244. Cifone, M.G., et al., Induction of the nitric oxide-synthesizing pathway in fresh and interleukin 2-cultured rat natural killer cells. Cell Immunol, 1994. 157(1): p. 181-94. 245. Juretic, A., et al., Generation of lymphokine-activated killer activity in rodents but not in humans is nitric oxide dependent. Cell Immunol, 1994. 157(2): p. 462-77.

179

246. Jyothi, M.D. and A. Khar, Interleukin-2-induced nitric oxide synthase and nuclear factor-kappaB activity in activated natural killer cells and the production of interferon-gamma. Scand J Immunol, 2000. 52(2): p. 148-55. 247. Orucevic, A. and P.K. Lala, Effects of N(G)-Nitro-L-arginine methyl ester, an inhibitor of nitric oxide synthesis, on IL-2-induced LAK cell generation in vivo and in vitro in healthy and tumor-bearing mice. Cell Immunol, 1996. 169(1): p. 125-32. 248. Diefenbach, A., et al., Requirement for type 2 NO synthase for IL-12 signaling in innate immunity. Science, 1999. 284(5416): p. 951-5. 249. Galli, S.J., New concepts about the mast cell. N Engl J Med, 1993. 328(4): p. 257-65. 250. Urb, M. and D.C. Sheppard, The role of mast cells in the defence against pathogens. PLoS Pathog, 2012. 8(4): p. e1002619. 251. Gaboury, J.P., X.F. Niu, and P. Kubes, Nitric oxide inhibits numerous features of mast cell-induced inflammation. Circulation, 1996. 93(2): p. 318-26. 252. Kimura, M., et al., Mast cell degranulation in rat mesenteric venule: effects of L-NAME, methylene blue and ketotifen. Pharmacol Res, 1999. 39(5): p. 397-402. 253. van Overveld, F.J.B., H.; Vermeire, P.A.; Herman, A.G., Nitroprusside, a nitrogen oxide generating drug, inhibits release of histamine and tryptase from human skin mast cells. Agents Actions, 1993. 38: p. C237-8. 254. Masini, E., et al., Rat mast cells synthesize a nitric oxide like-factor which modulates the release of histamine. Agents Actions, 1991. 33(1-2): p. 61-3. 255. Masini, E., et al., Nitric oxide modulates cardiac and mast cell anaphylaxis. Agents Actions, 1994. 41 Spec No: p. C89-90. 256. Kanwar, S., et al., Nitric oxide synthesis inhibition increases epithelial permeability via mast cells. Am J Physiol, 1994. 266(2 Pt 1): p. G222-9. 257. Bissonnette, E.Y., et al., Potentiation of tumor necrosis factor-alpha- mediated cytotoxicity of mast cells by their production of nitric oxide. J Immunol, 1991. 147(9): p. 3060-5. 258. Coleman, J.W., Nitric oxide: a regulator of mast cell activation and mast cell-mediated inflammation. Clin Exp Immunol, 2002. 129(1): p. 4-10. 259. Kobayashi, S.D., et al., Neutrophils in the innate immune response. Arch Immunol Ther Exp (Warsz), 2005. 53(6): p. 505-17. 260. Benjamim, C.F., S.H. Ferreira, and F.Q. Cunha, Role of nitric oxide in the failure of neutrophil migration in sepsis. J Infect Dis, 2000. 182(1): p. 214- 23. 261. Nolan, S., et al., Nitric oxide regulates neutrophil migration through microparticle formation. Am J Pathol, 2008. 172(1): p. 265-73. 262. Van Dervort, A.L., et al., Nitric oxide regulates endotoxin-induced TNF- alpha production by human neutrophils. J Immunol, 1994. 152(8): p. 4102- 9. 263. Ajuebor, M.N., et al., Role of inducible nitric oxide synthase in the regulation of neutrophil migration in zymosan-induced inflammation. Immunology, 1998. 95(4): p. 625-30. 264. Kobayashi, Y., The regulatory role of nitric oxide in proinflammatory cytokine expression during the induction and resolution of inflammation. J Leukoc Biol, 2010. 88(6): p. 1157-62.

180

265. Beauvais, F., L. Michel, and L. Dubertret, Exogenous nitric oxide elicits chemotaxis of neutrophils in vitro. J Cell Physiol, 1995. 165(3): p. 610-4. 266. Wanikiat, P., D.F. Woodward, and R.A. Armstrong, Investigation of the role of nitric oxide and cyclic GMP in both the activation and inhibition of human neutrophils. Br J Pharmacol, 1997. 122(6): p. 1135-45. 267. S. Ibiza, J.M.S., The role of nitric oxide in the regulation of adaptive immune responses. Revisión, 2008. 27(3): p. 103-117. 268. Mazzoni, A., et al., Myeloid suppressor lines inhibit T cell responses by an NO-dependent mechanism. J Immunol, 2002. 168(2): p. 689-95. 269. Niedbala, W., B. Cai, and F.Y. Liew, Role of nitric oxide in the regulation of T cell functions. Ann Rheum Dis, 2006. 65 Suppl 3: p. iii37-40. 270. Huang, F.P., et al., Nitric oxide regulates Th1 cell development through the inhibition of IL-12 synthesis by macrophages. Eur J Immunol, 1998. 28(12): p. 4062-70. 271. Wei, X.Q., et al., Altered immune responses in mice lacking inducible nitric oxide synthase. Nature, 1995. 375(6530): p. 408-11. 272. Tarrant, T.K., et al., Interleukin 12 protects from a T helper type 1-mediated autoimmune disease, experimental autoimmune uveitis, through a mechanism involving interferon gamma, nitric oxide, and apoptosis. J Exp Med, 1999. 189(2): p. 219-30. 273. Niedbala, W., et al., Nitric oxide preferentially induces type 1 T cell differentiation by selectively up-regulating IL-12 receptor beta 2 expression via cGMP. Proc Natl Acad Sci U S A, 2002. 99(25): p. 16186-91. 274. Okuda, Y., et al., Nitric oxide induces apoptosis in mouse splenic T lymphocytes. Immunol Lett, 1996. 52(2-3): p. 135-8. 275. Williams, M.S., et al., Nitric oxide synthase plays a signaling role in TCR- triggered apoptotic death. J Immunol, 1998. 161(12): p. 6526-31. 276. Li, C.Q., L.J. Trudel, and G.N. Wogan, Nitric oxide-induced genotoxicity, mitochondrial damage, and apoptosis in human lymphoblastoid cells expressing wild-type and mutant p53. Proc Natl Acad Sci U S A, 2002. 99(16): p. 10364-9. 277. Vig, M., et al., Inducible nitric oxide synthase in T cells regulates T cell death and immune memory. J Clin Invest, 2004. 113(12): p. 1734-42. 278. Ushmorov, A., et al., Nitric-oxide-induced apoptosis in human leukemic lines requires mitochondrial lipid degradation and cytochrome C release. Blood, 1999. 93(7): p. 2342-52. 279. Hara, M.R., et al., S-nitrosylated GAPDH initiates apoptotic cell death by nuclear translocation following Siah1 binding. Nat Cell Biol, 2005. 7(7): p. 665-74. 280. Dimmeler, S., et al., Nitric oxide inhibits APO-1/Fas-mediated cell death. Cell Growth Differ, 1998. 9(5): p. 415-22. 281. Mahidhara, R.S., et al., Nitric oxide-mediated inhibition of caspase- dependent T lymphocyte proliferation. J Leukoc Biol, 2003. 74(3): p. 403- 11. 282. Atochina, O., et al., A schistosome-expressed immunomodulatory glycoconjugate expands peritoneal Gr1(+) macrophages that suppress naive CD4(+) T cell proliferation via an IFN-gamma and nitric oxide-dependent mechanism. J Immunol, 2001. 167(8): p. 4293-302.

181

283. Aiello, S., et al., DnIKK2-transfected dendritic cells induce a novel population of inducible nitric oxide synthase-expressing CD4+CD25- cells with tolerogenic properties. Transplantation, 2007. 83(4): p. 474-84. 284. Niedbala, W., et al., Nitric oxide induces CD4+CD25+ Foxp3 regulatory T cells from CD4+CD25 T cells via p53, IL-2, and OX40. Proc Natl Acad Sci U S A, 2007. 104(39): p. 15478-83. 285. Taylor, M.W. and G.S. Feng, Relationship between interferon-gamma, indoleamine 2,3-dioxygenase, and tryptophan catabolism. FASEB J, 1991. 5(11): p. 2516-22. 286. Hassanain, H.H., S.Y. Chon, and S.L. Gupta, Differential regulation of human indoleamine 2,3-dioxygenase gene expression by interferons-gamma and - alpha. Analysis of the regulatory region of the gene and identification of an interferon-gamma-inducible DNA-binding factor. J Biol Chem, 1993. 268(7): p. 5077-84. 287. Belladonna, M.L., et al., Cutting edge: Autocrine TGF-beta sustains default tolerogenesis by IDO-competent dendritic cells. J Immunol, 2008. 181(8): p. 5194-8. 288. Pallotta, M.T., et al., Indoleamine 2,3-dioxygenase is a signaling protein in long-term tolerance by dendritic cells. Nat Immunol, 2011. 12(9): p. 870-8. 289. Grohmann, U., et al., CTLA-4-Ig regulates tryptophan catabolism in vivo. Nat Immunol, 2002. 3(11): p. 1097-101. 290. Fallarino, F., et al., Modulation of tryptophan catabolism by regulatory T cells. Nat Immunol, 2003. 4(12): p. 1206-12. 291. King, N.J. and S.R. Thomas, Molecules in focus: indoleamine 2,3- dioxygenase. Int J Biochem Cell Biol, 2007. 39(12): p. 2167-72. 292. Munn, D.H., et al., Inhibition of T cell proliferation by macrophage tryptophan catabolism. J Exp Med, 1999. 189(9): p. 1363-72. 293. Mellor, A.L. and D.H. Munn, Tryptophan catabolism and T-cell tolerance: immunosuppression by starvation? Immunol Today, 1999. 20(10): p. 469- 73. 294. Lee, G.K., et al., Tryptophan deprivation sensitizes activated T cells to apoptosis prior to cell division. Immunology, 2002. 107(4): p. 452-60. 295. Wek, R.C., H.Y. Jiang, and T.G. Anthony, Coping with stress: eIF2 kinases and translational control. Biochem Soc Trans, 2006. 34(Pt 1): p. 7-11. 296. Munn, D.H., et al., GCN2 kinase in T cells mediates proliferative arrest and anergy induction in response to indoleamine 2,3-dioxygenase. Immunity, 2005. 22(5): p. 633-42. 297. Fallarino, F., et al., T cell apoptosis by tryptophan catabolism. Cell Death Differ, 2002. 9(10): p. 1069-77. 298. Grohmann, U., F. Fallarino, and P. Puccetti, Tolerance, DCs and tryptophan: much ado about IDO. Trends Immunol, 2003. 24(5): p. 242-8. 299. Fernandez-Salguero, P.M., et al., Aryl-hydrocarbon receptor-deficient mice are resistant to 2,3,7,8-tetrachlorodibenzo-p-dioxin-induced toxicity. Toxicol Appl Pharmacol, 1996. 140(1): p. 173-9. 300. Mimura, J., et al., Loss of teratogenic response to 2,3,7,8-tetrachlorodibenzo- p-dioxin (TCDD) in mice lacking the Ah (dioxin) receptor. Genes Cells, 1997. 2(10): p. 645-54.

182

301. Puga, A., et al., Aromatic hydrocarbon receptor interaction with the retinoblastoma protein potentiates repression of E2F-dependent transcription and cell cycle arrest. J Biol Chem, 2000. 275(4): p. 2943-50. 302. Marlowe, J.L., et al., The aryl hydrocarbon receptor displaces p300 from E2F-dependent promoters and represses S phase-specific gene expression. J Biol Chem, 2004. 279(28): p. 29013-22. 303. Marlowe, J.L., et al., The aryl hydrocarbon receptor binds to E2F1 and inhibits E2F1-induced apoptosis. Mol Biol Cell, 2008. 19(8): p. 3263-71. 304. Veldhoen, M., et al., The aryl hydrocarbon receptor links TH17-cell- mediated autoimmunity to environmental toxins. Nature, 2008. 453(7191): p. 106-9. 305. Quintana, F.J., et al., Control of T(reg) and T(H)17 cell differentiation by the aryl hydrocarbon receptor. Nature, 2008. 453(7191): p. 65-71. 306. Platten, M., W. Wick, and B.J. Van den Eynde, Tryptophan catabolism in cancer: beyond IDO and tryptophan depletion. Cancer Res, 2012. 72(21): p. 5435-40. 307. Fallarino, F., et al., The combined effects of tryptophan starvation and tryptophan catabolites down-regulate T cell receptor zeta-chain and induce a regulatory phenotype in naive T cells. J Immunol, 2006. 176(11): p. 6752-61. 308. Grohmann, U., et al., A defect in tryptophan catabolism impairs tolerance in nonobese diabetic mice. J Exp Med, 2003. 198(1): p. 153-60. 309. Uyttenhove, C., et al., Evidence for a tumoral immune resistance mechanism based on tryptophan degradation by indoleamine 2,3-dioxygenase. Nat Med, 2003. 9(10): p. 1269-74. 310. Jasperson, L.K., et al., Indoleamine 2,3-dioxygenase is a critical regulator of acute graft-versus-host disease lethality. Blood, 2008. 111(6): p. 3257-65. 311. Park, G., et al., A paradoxical pattern of indoleamine 2,3-dioxygenase expression in the colon tissues of patients with acute graft-versus-host disease. Exp Hematol, 2014. 42(9): p. 734-40. 312. Yan, Y., et al., IDO upregulates regulatory T cells via tryptophan catabolite and suppresses encephalitogenic T cell responses in experimental autoimmune encephalomyelitis. J Immunol, 2010. 185(10): p. 5953-61. 313. Williams, R.O., Exploitation of the IDO Pathway in the Therapy of Rheumatoid Arthritis. Int J Tryptophan Res, 2013. 6(Suppl 1): p. 67-73. 314. Zhu, L., et al., Synovial autoreactive T cells in rheumatoid arthritis resist IDO-mediated inhibition. J Immunol, 2006. 177(11): p. 8226-33. 315. Munn, D.H. and A.L. Mellor, Indoleamine 2,3-dioxygenase and tumor- induced tolerance. J Clin Invest, 2007. 117(5): p. 1147-54. 316. Balachandran, V.P., et al., Imatinib potentiates antitumor T cell responses in gastrointestinal stromal tumor through the inhibition of Ido. Nat Med, 2011. 17(9): p. 1094-100. 317. Schneider, E., et al., Histamine-producing cell-stimulating activity. Interleukin 3 and granulocyte-macrophage colony-stimulating factor induce de novo synthesis of histidine decarboxylase in hemopoietic progenitor cells. J Immunol, 1987. 139(11): p. 3710-7. 318. Kuramasu, A., et al., Mast cell-/basophil-specific transcriptional regulation of human L-histidine decarboxylase gene by CpG methylation in the promoter region. J Biol Chem, 1998. 273(47): p. 31607-14.

183

319. Shiraishi, M., et al., Participation of mitogen-activated protein kinase in thapsigargin- and TPA-induced histamine production in murine macrophage RAW 264.7 cells. Br J Pharmacol, 2000. 129(3): p. 515-24. 320. Szeberenyi, J.B., et al., Inhibition of effects of endogenously synthesized histamine disturbs in vitro human dendritic cell differentiation. Immunol Lett, 2001. 76(3): p. 175-82. 321. Shiraishi, M., et al., Analysis of histamine-producing cells at the late phase of allergic inflammation in rats. Immunology, 2000. 99(4): p. 600-6. 322. Aoi, R., et al., Histamine synthesis by mouse T lymphocytes through induced histidine decarboxylase. Immunology, 1989. 66(2): p. 219-23. 323. Jutel, M., et al., Histamine regulates T-cell and antibody responses by differential expression of H1 and H2 receptors. Nature, 2001. 413(6854): p. 420-5. 324. Dunford, P.J., et al., The histamine H4 receptor mediates allergic airway inflammation by regulating the activation of CD4+ T cells. J Immunol, 2006. 176(11): p. 7062-70. 325. Koncz, A., et al., Nitric oxide mediates T cell cytokine production and signal transduction in histidine decarboxylase knockout mice. J Immunol, 2007. 179(10): p. 6613-9. 326. Lanzavecchia, A., G. Iezzi, and A. Viola, From TCR engagement to T cell activation: a kinetic view of T cell behavior. Cell, 1999. 96(1): p. 1-4. 327. Simon, T., V. Laszlo, and A. Falus, Impact of histamine on dendritic cell functions. Cell Biol Int, 2011. 35(10): p. 997-1000. 328. Caron, G., et al., Histamine polarizes human dendritic cells into Th2 cell- promoting effector dendritic cells. J Immunol, 2001. 167(7): p. 3682-6. 329. Mazzoni, A., et al., Histamine regulates cytokine production in maturing dendritic cells, resulting in altered T cell polarization. J Clin Invest, 2001. 108(12): p. 1865-73. 330. Pavlinkova, G., et al., Effects of histamine on functional maturation of dendritic cells. Immunobiology, 2003. 207(5): p. 315-25. 331. Gutzmer, R., et al., Histamine H4 receptor stimulation suppresses IL-12p70 production and mediates chemotaxis in human monocyte-derived dendritic cells. J Immunol, 2005. 174(9): p. 5224-32. 332. Lock, C., et al., Gene-microarray analysis of multiple sclerosis lesions yields new targets validated in autoimmune encephalomyelitis. Nat Med, 2002. 8(5): p. 500-8. 333. Ma, R.Z., et al., Identification of Bphs, an autoimmune disease locus, as histamine receptor H1. Science, 2002. 297(5581): p. 620-3. 334. Alonso, A., S.S. Jick, and M.A. Hernan, Allergy, histamine 1 receptor blockers, and the risk of multiple sclerosis. Neurology, 2006. 66(4): p. 572- 5. 335. Logothetis, L., et al., A pilot, open label, clinical trial using hydroxyzine in multiple sclerosis. Int J Immunopathol Pharmacol, 2005. 18(4): p. 771-8. 336. Black, J.W., et al., Definition and antagonism of histamine H 2 -receptors. Nature, 1972. 236(5347): p. 385-90. 337. Rajasekaran, N., et al., Histidine decarboxylase but not histamine receptor 1 or 2 deficiency protects from K/BxN serum-induced arthritis. Int Immunol, 2009. 21(11): p. 1263-8.

184

338. Cowden, J.M., et al., Histamine H4 receptor antagonism diminishes existing airway inflammation and dysfunction via modulation of Th2 cytokines. Respir Res, 2010. 11: p. 86. 339. Cowden, J.M., et al., The histamine H4 receptor mediates inflammation and pruritus in Th2-dependent dermal inflammation. J Invest Dermatol, 2010. 130(4): p. 1023-33. 340. Beermann, S., et al., Opposite effects of mepyramine on JNJ 7777120- induced amelioration of experimentally induced asthma in mice in sensitization and provocation. PLoS One, 2012. 7(1): p. e30285. 341. Takeshita, K., et al., Critical role of histamine H4 receptor in leukotriene B4 production and mast cell-dependent neutrophil recruitment induced by zymosan in vivo. J Pharmacol Exp Ther, 2003. 307(3): p. 1072-8. 342. del Rio, R., et al., Histamine H4 receptor optimizes T regulatory cell frequency and facilitates anti-inflammatory responses within the central nervous system. J Immunol, 2012. 188(2): p. 541-7. 343. Morgan, R.K., et al., Histamine 4 receptor activation induces recruitment of FoxP3+ T cells and inhibits allergic asthma in a murine model. J Immunol, 2007. 178(12): p. 8081-9. 344. Boulland, M.L., et al., Human IL4I1 is a secreted L-phenylalanine oxidase expressed by mature dendritic cells that inhibits T-lymphocyte proliferation. Blood, 2007. 110(1): p. 220-7. 345. Chavan, S.S., et al., Characterization of the human homolog of the IL-4 induced gene-1 (Fig1). Biochim Biophys Acta, 2002. 1576(1-2): p. 70-80. 346. Copie-Bergman, C., et al., Interleukin 4-induced gene 1 is activated in primary mediastinal large B-cell lymphoma. Blood, 2003. 101(7): p. 2756- 61. 347. Chu, C.C. and W.E. Paul, Fig1, an interleukin 4-induced mouse B cell gene isolated by cDNA representational difference analysis. Proc Natl Acad Sci U S A, 1997. 94(6): p. 2507-12. 348. Marquet, J., et al., Dichotomy between factors inducing the immunosuppressive enzyme IL-4-induced gene 1 (IL4I1) in B lymphocytes and mononuclear phagocytes. Eur J Immunol, 2010. 40(9): p. 2557-68. 349. Lasoudris, F., et al., IL4I1: an inhibitor of the CD8(+) antitumor T-cell response in vivo. Eur J Immunol, 2011. 41(6): p. 1629-38. 350. Puiffe, M.L., et al., Antibacterial properties of the mammalian L-amino acid oxidase IL4I1. PLoS One, 2013. 8(1): p. e54589. 351. Santarlasci, V., et al., Rarity of human T helper 17 cells is due to retinoic acid orphan receptor-dependent mechanisms that limit their expansion. Immunity, 2012. 36(2): p. 201-14. 352. Trumpp, A., M. Essers, and A. Wilson, Awakening dormant haematopoietic stem cells. Nat Rev Immunol, 2010. 10(3): p. 201-9. 353. Wilson, A. and A. Trumpp, Bone-marrow haematopoietic-stem-cell niches. Nat Rev Immunol, 2006. 6(2): p. 93-106. 354. Lo Celso, C. and D.T. Scadden, The haematopoietic stem cell niche at a glance. J Cell Sci, 2011. 124(Pt 21): p. 3529-35. 355. Dykstra, B., et al., Long-term propagation of distinct hematopoietic differentiation programs in vivo. Cell Stem Cell, 2007. 1(2): p. 218-29. 356. Lo Celso, C., et al., Live-animal tracking of individual haematopoietic stem/progenitor cells in their niche. Nature, 2009. 457(7225): p. 92-6.

185

357. Wilson, A., et al., Hematopoietic stem cells reversibly switch from dormancy to self-renewal during homeostasis and repair. Cell, 2008. 135(6): p. 1118- 29. 358. Kiel, M.J. and S.J. Morrison, Uncertainty in the niches that maintain haematopoietic stem cells. Nat Rev Immunol, 2008. 8(4): p. 290-301. 359. El-Badri, N.S., et al., Osteoblasts promote engraftment of allogeneic hematopoietic stem cells. Exp Hematol, 1998. 26(2): p. 110-6. 360. Ellis, S.L., et al., The relationship between bone, hemopoietic stem cells, and vasculature. Blood, 2011. 118(6): p. 1516-24. 361. Calvi, L.M., et al., Osteoblastic cells regulate the haematopoietic stem cell niche. Nature, 2003. 425(6960): p. 841-6. 362. Zhang, J., et al., Identification of the haematopoietic stem cell niche and control of the niche size. Nature, 2003. 425(6960): p. 836-41. 363. Lymperi, S., et al., Strontium can increase some osteoblasts without increasing hematopoietic stem cells. Blood, 2008. 111(3): p. 1173-81. 364. Kiel, M.J., G.L. Radice, and S.J. Morrison, Lack of evidence that hematopoietic stem cells depend on N-cadherin-mediated adhesion to osteoblasts for their maintenance. Cell Stem Cell, 2007. 1(2): p. 204-17. 365. Kiel, M.J., et al., SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell, 2005. 121(7): p. 1109-21. 366. Kiel, M.J., et al., Hematopoietic stem cells do not depend on N-cadherin to regulate their maintenance. Cell Stem Cell, 2009. 4(2): p. 170-9. 367. Greenbaum, A.M., et al., N-cadherin in osteolineage cells is not required for maintenance of hematopoietic stem cells. Blood, 2012. 120(2): p. 295-302. 368. Bromberg, O., et al., Osteoblastic N-cadherin is not required for microenvironmental support and regulation of hematopoietic stem and progenitor cells. Blood, 2012. 120(2): p. 303-13. 369. Taichman, R.S., M.J. Reilly, and S.G. Emerson, Human osteoblasts support human hematopoietic progenitor cells in vitro bone marrow cultures. Blood, 1996. 87(2): p. 518-24. 370. Arai, F., et al., Tie2/angiopoietin-1 signaling regulates hematopoietic stem cell quiescence in the bone marrow niche. Cell, 2004. 118(2): p. 149-61. 371. Goldman, D.C., et al., BMP4 regulates the hematopoietic stem cell niche. Blood, 2009. 114(20): p. 4393-401. 372. Stier, S., et al., Notch1 activation increases hematopoietic stem cell self- renewal in vivo and favors lymphoid over myeloid lineage outcome. Blood, 2002. 99(7): p. 2369-78. 373. Karanu, F.N., et al., The notch ligand jagged-1 represents a novel growth factor of human hematopoietic stem cells. J Exp Med, 2000. 192(9): p. 1365-72. 374. Mancini, S.J., et al., Jagged1-dependent Notch signaling is dispensable for hematopoietic stem cell self-renewal and differentiation. Blood, 2005. 105(6): p. 2340-2. 375. Gekas, C., et al., The placenta is a niche for hematopoietic stem cells. Dev Cell, 2005. 8(3): p. 365-75. 376. Ohneda, O., et al., Hematopoietic stem cell maintenance and differentiation are supported by embryonic aorta-gonad-mesonephros region-derived endothelium. Blood, 1998. 92(3): p. 908-19.

186

377. Li, W., et al., Primary endothelial cells isolated from the yolk sac and para- aortic splanchnopleura support the expansion of adult marrow stem cells in vitro. Blood, 2003. 102(13): p. 4345-53. 378. Hooper, A.T., et al., Engraftment and reconstitution of hematopoiesis is dependent on VEGFR2-mediated regeneration of sinusoidal endothelial cells. Cell Stem Cell, 2009. 4(3): p. 263-74. 379. Kobayashi, H., et al., Angiocrine factors from Akt-activated endothelial cells balance self-renewal and differentiation of haematopoietic stem cells. Nat Cell Biol, 2010. 12(11): p. 1046-56. 380. Schweitzer, K.M., et al., Constitutive expression of E-selectin and vascular cell adhesion molecule-1 on endothelial cells of hematopoietic tissues. Am J Pathol, 1996. 148(1): p. 165-75. 381. Avecilla, S.T., et al., Chemokine-mediated interaction of hematopoietic progenitors with the bone marrow vascular niche is required for thrombopoiesis. Nat Med, 2004. 10(1): p. 64-71. 382. Bruns, I., et al., Megakaryocytes regulate hematopoietic stem cell quiescence through CXCL4 secretion. Nat Med, 2014. 20(11): p. 1315-20. 383. Zhao, M., et al., Megakaryocytes maintain homeostatic quiescence and promote post-injury regeneration of hematopoietic stem cells. Nat Med, 2014. 20(11): p. 1321-6. 384. Li, J.J., et al., Thrombin induces the release of angiopoietin-1 from platelets. Thromb Haemost, 2001. 85(2): p. 204-6. 385. Sugiyama, T., et al., Maintenance of the hematopoietic stem cell pool by CXCL12-CXCR4 chemokine signaling in bone marrow stromal cell niches. Immunity, 2006. 25(6): p. 977-88. 386. Pinho, S., et al., PDGFRalpha and CD51 mark human nestin+ sphere-forming mesenchymal stem cells capable of hematopoietic progenitor cell expansion. J Exp Med, 2013. 210(7): p. 1351-67. 387. Dexter, T.M., T.D. Allen, and L.G. Lajtha, Conditions controlling the proliferation of haemopoietic stem cells in vitro. J Cell Physiol, 1977. 91(3): p. 335-44. 388. Naveiras, O., et al., Bone-marrow adipocytes as negative regulators of the haematopoietic microenvironment. Nature, 2009. 460(7252): p. 259-63. 389. Haynesworth, S.E., M.A. Baber, and A.I. Caplan, Cytokine expression by human marrow-derived mesenchymal progenitor cells in vitro: effects of dexamethasone and IL-1 alpha. J Cell Physiol, 1996. 166(3): p. 585-92. 390. Majumdar, M.K., et al., Phenotypic and functional comparison of cultures of marrow-derived mesenchymal stem cells (MSCs) and stromal cells. J Cell Physiol, 1998. 176(1): p. 57-66. 391. Ikebuchi, K., et al., Interleukin 6 enhancement of interleukin 3-dependent proliferation of multipotential hemopoietic progenitors. Proc Natl Acad Sci U S A, 1987. 84(24): p. 9035-9. 392. Bordeaux-Rego, P., et al., Both interleukin-3 and interleukin-6 are necessary for better ex vivo expansion of CD133+ cells from umbilical cord blood. Stem Cells Dev, 2010. 19(3): p. 413-22. 393. Majumdar, M.K., et al., Human marrow-derived mesenchymal stem cells (MSCs) express hematopoietic cytokines and support long-term hematopoiesis when differentiated toward stromal and osteogenic lineages. J Hematother Stem Cell Res, 2000. 9(6): p. 841-8.

187

394. Escary, J.L., et al., Leukaemia inhibitory factor is necessary for maintenance of haematopoietic stem cells and thymocyte stimulation. Nature, 1993. 363(6427): p. 361-4. 395. Broxmeyer, H.E., et al., Flt3 ligand stimulates/costimulates the growth of myeloid stem/progenitor cells. Exp Hematol, 1995. 23(10): p. 1121-9. 396. Petzer, A.L., et al., Differential cytokine effects on primitive (CD34+CD38-) human hematopoietic cells: novel responses to Flt3-ligand and thrombopoietin. J Exp Med, 1996. 183(6): p. 2551-8. 397. Ducos, K., et al., p21(cip1) mRNA is controlled by endogenous transforming growth factor-beta1 in quiescent human hematopoietic stem/progenitor cells. J Cell Physiol, 2000. 184(1): p. 80-5. 398. Mendez-Ferrer, S., et al., Mesenchymal and haematopoietic stem cells form a unique bone marrow niche. Nature, 2010. 466(7308): p. 829-34. 399. Katayama, Y., et al., Signals from the sympathetic nervous system regulate hematopoietic stem cell egress from bone marrow. Cell, 2006. 124(2): p. 407-21. 400. Isern, J., et al., The neural crest is a source of mesenchymal stem cells with specialized hematopoietic stem cell niche function. Elife, 2014. 3: p. e03696. 401. Omatsu, Y., et al., The essential functions of adipo-osteogenic progenitors as the hematopoietic stem and progenitor cell niche. Immunity, 2010. 33(3): p. 387-99. 402. Park, D., et al., Endogenous bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and regeneration. Cell Stem Cell, 2012. 10(3): p. 259-72. 403. Ding, L., et al., Endothelial and perivascular cells maintain haematopoietic stem cells. Nature, 2012. 481(7382): p. 457-62. 404. Zhou, B.O., et al., Leptin-receptor-expressing mesenchymal stromal cells represent the main source of bone formed by adult bone marrow. Cell Stem Cell, 2014. 15(2): p. 154-68. 405. Ding, L. and S.J. Morrison, Haematopoietic stem cells and early lymphoid progenitors occupy distinct bone marrow niches. Nature, 2013. 495(7440): p. 231-5. 406. Kunisaki, Y., et al., Arteriolar niches maintain haematopoietic stem cell quiescence. Nature, 2013. 502(7473): p. 637-43. 407. Mercier, F.E., C. Ragu, and D.T. Scadden, The bone marrow at the crossroads of blood and immunity. Nat Rev Immunol, 2012. 12(1): p. 49- 60. 408. Isern, J. and S. Mendez-Ferrer, Stem cell interactions in a bone marrow niche. Curr Osteoporos Rep, 2011. 9(4): p. 210-8. 409. Noort, W.A., et al., Mesenchymal stem cells promote engraftment of human umbilical cord blood-derived CD34(+) cells in NOD/SCID mice. Exp Hematol, 2002. 30(8): p. 870-8. 410. Zhang, Y., et al., Simultaneous injection of bone marrow cells and stromal cells into bone marrow accelerates hematopoiesis in vivo. Stem Cells, 2004. 22(7): p. 1256-62. 411. Han, J.Y., et al., Cotransplantation of cord blood hematopoietic stem cells and culture-expanded and GM-CSF-/SCF-transfected mesenchymal stem cells in SCID mice. J Korean Med Sci, 2007. 22(2): p. 242-7.

188

412. Sugiura, K., et al., Major histocompatibility complex restriction between hematopoietic stem cells and stromal cells in vitro. Stem Cells, 2001. 19(1): p. 46-58. 413. Koc, O.N., et al., Rapid hematopoietic recovery after coinfusion of autologous-blood stem cells and culture-expanded marrow mesenchymal stem cells in advanced breast cancer patients receiving high-dose chemotherapy. J Clin Oncol, 2000. 18(2): p. 307-16. 414. Lazarus, H.M., et al., Cotransplantation of HLA-identical sibling culture- expanded mesenchymal stem cells and hematopoietic stem cells in hematologic malignancy patients. Biol Blood Marrow Transplant, 2005. 11(5): p. 389-98. 415. Gurevitch, O., et al., Transplantation of allogeneic or xenogeneic bone marrow within the donor stromal microenvironment. Transplantation, 1999. 68(9): p. 1362-8. 416. Tian, Y., et al., Bone marrow-derived mesenchymal stem cells decrease acute graft-versus-host disease after allogeneic hematopoietic stem cells transplantation. Immunol Invest, 2008. 37(1): p. 29-42. 417. Le Blanc, K., et al., Mesenchymal stem cells for treatment of steroid- resistant, severe, acute graft-versus-host disease: a phase II study. Lancet, 2008. 371(9624): p. 1579-86. 418. Sudres, M., et al., Bone marrow mesenchymal stem cells suppress lymphocyte proliferation in vitro but fail to prevent graft-versus-host disease in mice. J Immunol, 2006. 176(12): p. 7761-7. 419. Tisato, V., et al., Mesenchymal stem cells of cord blood origin are effective at preventing but not treating graft-versus-host disease. Leukemia, 2007. 21(9): p. 1992-9. 420. Polchert, D., et al., IFN-gamma activation of mesenchymal stem cells for treatment and prevention of graft versus host disease. Eur J Immunol, 2008. 38(6): p. 1745-55. 421. Wang, Y., et al., The Wnt/beta-catenin pathway is required for the development of leukemia stem cells in AML. Science, 2010. 327(5973): p. 1650-3. 422. Hoang, V.T., A. Zepeda-Moreno, and A.D. Ho, Identification of leukemia stem cells in acute myeloid leukemia and their clinical relevance. Biotechnol J, 2012. 7(6): p. 779-88. 423. Ashton, J.M., et al., Gene sets identified with oncogene cooperativity analysis regulate in vivo growth and survival of leukemia stem cells. Cell Stem Cell, 2012. 11(3): p. 359-72. 424. Colmone, A., et al., Leukemic cells create bone marrow niches that disrupt the behavior of normal hematopoietic progenitor cells. Science, 2008. 322(5909): p. 1861-5. 425. Konopleva, M.Y. and C.T. Jordan, Leukemia stem cells and microenvironment: biology and therapeutic targeting. J Clin Oncol, 2011. 29(5): p. 591-9. 426. Nair, R.R., J. Tolentino, and L.A. Hazlehurst, The bone marrow microenvironment as a sanctuary for minimal residual disease in CML. Biochem Pharmacol, 2010. 80(5): p. 602-12.

189

427. Guan, Y., B. Gerhard, and D.E. Hogge, Detection, isolation, and stimulation of quiescent primitive leukemic progenitor cells from patients with acute myeloid leukemia (AML). Blood, 2003. 101(8): p. 3142-9. 428. Holyoake, T.L., et al., Elucidating critical mechanisms of deregulated stem cell turnover in the chronic phase of chronic myeloid leukemia. Leukemia, 2002. 16(4): p. 549-58. 429. Holyoake, T.L., et al., Primitive quiescent leukemic cells from patients with chronic myeloid leukemia spontaneously initiate factor-independent growth in vitro in association with up-regulation of expression of interleukin-3. Blood, 2001. 97(3): p. 720-8. 430. Shen, W., et al., The chemokine receptor CXCR4 enhances integrin-mediated in vitro adhesion and facilitates engraftment of leukemic precursor-B cells in the bone marrow. Exp Hematol, 2001. 29(12): p. 1439-47. 431. Raaijmakers, M.H., et al., Bone progenitor dysfunction induces myelodysplasia and secondary leukaemia. Nature, 2010. 464(7290): p. 852-7. 432. Rombouts, E.J., et al., Relation between CXCR-4 expression, Flt3 mutations, and unfavorable prognosis of adult acute myeloid leukemia. Blood, 2004. 104(2): p. 550-7. 433. Flores-Figueroa, E., et al., Distinctive contact between CD34+ hematopoietic progenitors and CXCL12+ CD271+ mesenchymal stromal cells in benign and myelodysplastic bone marrow. Lab Invest, 2012. 92(9): p. 1330-41. 434. Zeng, Z., et al., Inhibition of CXCR4 with the novel RCP168 peptide overcomes stroma-mediated chemoresistance in chronic and acute leukemias. Mol Cancer Ther, 2006. 5(12): p. 3113-21. 435. Tavor, S., et al., CXCR4 regulates migration and development of human acute myelogenous leukemia stem cells in transplanted NOD/SCID mice. Cancer Res, 2004. 64(8): p. 2817-24. 436. Vianello, F., et al., Bone marrow mesenchymal stromal cells non-selectively protect chronic myeloid leukemia cells from imatinib-induced apoptosis via the CXCR4/CXCL12 axis. Haematologica, 2010. 95(7): p. 1081-9. 437. Xu, Y., et al., TGF-beta receptor kinase inhibitor LY2109761 reverses the anti-apoptotic effects of TGF-beta1 in myelo-monocytic leukaemic cells co- cultured with stromal cells. Br J Haematol, 2008. 142(2): p. 192-201. 438. Tabe, Y., et al., TGF-beta-Neutralizing Antibody 1D11 Enhances Cytarabine- Induced Apoptosis in AML Cells in the Bone Marrow Microenvironment. PLoS One, 2013. 8(6): p. e62785. 439. Ghannam, S., et al., Immunosuppression by mesenchymal stem cells: mechanisms and clinical applications. Stem Cell Res Ther, 2010. 1(1): p. 2. 440. Bian, L., et al., In vitro and in vivo immunosuppressive characteristics of hepatocyte growth factor-modified murine mesenchymal stem cells. In Vivo, 2009. 23(1): p. 21-7. 441. Choi, J.J., et al., Mesenchymal stem cells overexpressing interleukin-10 attenuate collagen-induced arthritis in mice. Clin Exp Immunol, 2008. 153(2): p. 269-76. 442. Nasef, A., et al., Immunosuppressive effects of mesenchymal stem cells: involvement of HLA-G. Transplantation, 2007. 84(2): p. 231-7. 443. Kota, D.J., et al., Differential MSC activation leads to distinct mononuclear leukocyte binding mechanisms. Sci Rep, 2014. 4: p. 4565.

190

444. Mastri, M., et al., Activation of Toll-like receptor 3 amplifies mesenchymal stem cell trophic factors and enhances therapeutic potency. Am J Physiol Cell Physiol, 2012. 303(10): p. C1021-33. 445. Sonoki, T., et al., Coinduction of nitric-oxide synthase and arginase I in cultured rat peritoneal macrophages and rat tissues in vivo by lipopolysaccharide. J Biol Chem, 1997. 272(6): p. 3689-93. 446. Chen, X., J.J. Oppenheim, and O.M. Howard, BALB/c mice have more CD4+CD25+ T regulatory cells and show greater susceptibility to suppression of their CD4+CD25- responder T cells than C57BL/6 mice. J Leukoc Biol, 2005. 78(1): p. 114-21. 447. Laurent, M., M. Lepoivre, and J.P. Tenu, Kinetic modelling of the nitric oxide gradient generated in vitro by adherent cells expressing inducible nitric oxide synthase. Biochem J, 1996. 314 ( Pt 1): p. 109-13. 448. Albina, J.E., J.A. Abate, and W.L. Henry, Jr., Nitric oxide production is required for murine resident peritoneal macrophages to suppress mitogen- stimulated T cell proliferation. Role of IFN-gamma in the induction of the nitric oxide-synthesizing pathway. J Immunol, 1991. 147(1): p. 144-8. 449. Bonizzi, G. and M. Karin, The two NF-kappaB activation pathways and their role in innate and adaptive immunity. Trends Immunol, 2004. 25(6): p. 280-8. 450. Oeckinghaus, A., M.S. Hayden, and S. Ghosh, Crosstalk in NF-kappaB signaling pathways. Nat Immunol, 2011. 12(8): p. 695-708. 451. Hayden, M.S. and S. Ghosh, Signaling to NF-kappaB. Genes Dev, 2004. 18(18): p. 2195-224. 452. Honda, K., A. Takaoka, and T. Taniguchi, Type I interferon [corrected] gene induction by the interferon regulatory factor family of transcription factors. Immunity, 2006. 25(3): p. 349-60. 453. Kaisho, T. and S. Akira, Dendritic-cell function in Toll-like receptor- and MyD88-knockout mice. Trends Immunol, 2001. 22(2): p. 78-83. 454. Adachi, O., et al., Targeted disruption of the MyD88 gene results in loss of IL- 1- and IL-18-mediated function. Immunity, 1998. 9(1): p. 143-50. 455. Kawai, T., et al., Unresponsiveness of MyD88-deficient mice to endotoxin. Immunity, 1999. 11(1): p. 115-22. 456. Zhu, X., et al., MyD88 and NOS2 are essential for toll-like receptor 4- mediated survival effect in cardiomyocytes. Am J Physiol Heart Circ Physiol, 2006. 291(4): p. H1900-9. 457. Lima-Junior, D.S., et al., Inflammasome-derived IL-1beta production induces nitric oxide-mediated resistance to Leishmania. Nat Med, 2013. 19(7): p. 909-15. 458. Morikawa, S., et al., Development of mesenchymal stem cells partially originate from the neural crest. Biochem Biophys Res Commun, 2009. 379(4): p. 1114-9. 459. Chabannes, D., et al., A role for heme oxygenase-1 in the immunosuppressive effect of adult rat and human mesenchymal stem cells. Blood, 2007. 110(10): p. 3691-4. 460. Bartholomew, A., et al., Baboon mesenchymal stem cells can be genetically modified to secrete human erythropoietin in vivo. Hum Gene Ther, 2001. 12(12): p. 1527-41.

191

461. Nauta, A.J., et al., Mesenchymal stem cells inhibit generation and function of both CD34+-derived and monocyte-derived dendritic cells. J Immunol, 2006. 177(4): p. 2080-7. 462. Norian, L.A., et al., Tumor-infiltrating regulatory dendritic cells inhibit CD8+ T cell function via L-arginine metabolism. Cancer Res, 2009. 69(7): p. 3086-94. 463. Pesce, J.T., et al., Arginase-1-expressing macrophages suppress Th2 cytokine-driven inflammation and fibrosis. PLoS Pathog, 2009. 5(4): p. e1000371. 464. Ming, X.F., et al., Arginase II Promotes Macrophage Inflammatory Responses Through Mitochondrial Reactive Oxygen Species, Contributing to Insulin Resistance and Atherogenesis. J Am Heart Assoc, 2012. 1(4): p. e000992. 465. Bogdan, C., Nitric oxide and the immune response. Nat Immunol, 2001. 2(10): p. 907-16. 466. Galea, E. and D.L. Feinstein, Regulation of the expression of the inflammatory nitric oxide synthase (NOS2) by cyclic AMP. FASEB J, 1999. 13(15): p. 2125-37. 467. Shultz, P.J., S.L. Archer, and M.E. Rosenberg, Inducible nitric oxide synthase mRNA and activity in glomerular mesangial cells. Kidney Int, 1994. 46(3): p. 683-9. 468. Wang, R., et al., Human dermal fibroblasts produce nitric oxide and express both constitutive and inducible nitric oxide synthase isoforms. J Invest Dermatol, 1996. 106(3): p. 419-27. 469. Afzal, F., J. Polak, and L. Buttery, Endothelial nitric oxide synthase in the control of osteoblastic mineralizing activity and bone integrity. J Pathol, 2004. 202(4): p. 503-10. 470. Dimmeler, S., et al., Activation of nitric oxide synthase in endothelial cells by Akt-dependent phosphorylation. Nature, 1999. 399(6736): p. 601-5. 471. Ikeda, U., Y. Maeda, and K. Shimada, Inducible nitric oxide synthase and atherosclerosis. Clin Cardiol, 1998. 21(7): p. 473-6. 472. Hatano, E., et al., NF-kappaB stimulates inducible nitric oxide synthase to protect mouse hepatocytes from TNF-alpha- and Fas-mediated apoptosis. Gastroenterology, 2001. 120(5): p. 1251-62. 473. Shahid M, Tripathi T, and Sobial F, Histamine, Histamine Receptors, and their Role in Immunomodulation: An Updated Systematic Review. The Open Immunology Journal, 2009. 2: p. 9-41. 474. Laszlo, V. and A. Falus, Yet unrevealed aspects of histamine: questions from outside--answers at inside? Semin Cancer Biol, 2000. 10(1): p. 1-2. 475. Pochampally, R.R., et al., Histamine receptor H1 and dermatopontin: new downstream targets of the vitamin D receptor. J Bone Miner Res, 2007. 22(9): p. 1338-49. 476. Nemeth, K., et al., Characterization and Function of Histamine Receptors in Human Bone Marrow Stromal Cells. Stem Cells, 2011. 477. Munn, D.H., et al., Prevention of allogeneic fetal rejection by tryptophan catabolism. Science, 1998. 281(5380): p. 1191-3. 478. Xu, H., et al., Indoleamine 2,3-dioxygenase in lung dendritic cells promotes Th2 responses and allergic inflammation. Proc Natl Acad Sci U S A, 2008. 105(18): p. 6690-5.

192

479. Ling, W., et al., Mesenchymal stem cells use IDO to regulate immunity in tumor microenvironment. Cancer Res, 2014. 74(5): p. 1576-87. 480. Su, J., et al., Phylogenetic distinction of iNOS and IDO function in mesenchymal stem cell-mediated immunosuppression in mammalian species. Cell Death Differ, 2014. 21(3): p. 388-96. 481. Carbonnelle-Puscian, A., et al., The novel immunosuppressive enzyme IL4I1 is expressed by neoplastic cells of several B-cell lymphomas and by tumor- associated macrophages. Leukemia, 2009. 23(5): p. 952-60. 482. Schneider, E., et al., Trends in histamine research: new functions during immune responses and hematopoiesis. Trends Immunol, 2002. 23(5): p. 255-63. 483. Jutel, M., et al., Immune regulation by histamine. Curr Opin Immunol, 2002. 14(6): p. 735-40. 484. Akdis, C.A. and K. Blaser, Histamine in the immune regulation of allergic inflammation. J Allergy Clin Immunol, 2003. 112(1): p. 15-22. 485. Osna, N., et al., Histamine utilizes JAK-STAT pathway in regulating cytokine production. Int Immunopharmacol, 2001. 1(4): p. 759-62. 486. Lapilla, M., et al., Histamine regulates autoreactive T cell activation and adhesiveness in inflamed brain microcirculation. J Leukoc Biol, 2011. 89(2): p. 259-67. 487. Nemeth, K., et al., Characterization and function of histamine receptors in human bone marrow stromal cells. Stem Cells, 2012. 30(2): p. 222-31. 488. Lucas, R., et al., Arginase 1: an unexpected mediator of pulmonary capillary barrier dysfunction in models of acute lung injury. Front Immunol, 2013. 4: p. 228. 489. Salimuddin, et al., Regulation of the genes for arginase isoforms and related enzymes in mouse macrophages by lipopolysaccharide. Am J Physiol, 1999. 277(1 Pt 1): p. E110-7. 490. Oh, I., et al., Interferon-gamma and NF-kappaB mediate nitric oxide production by mesenchymal stromal cells. Biochem Biophys Res Commun, 2007. 355(4): p. 956-62. 491. Zinocker, S. and J.T. Vaage, Rat mesenchymal stromal cells inhibit T cell proliferation but not cytokine production through inducible nitric oxide synthase. Front Immunol, 2012. 3: p. 62. 492. Marks-Konczalik, J., S.C. Chu, and J. Moss, Cytokine-mediated transcriptional induction of the human inducible nitric oxide synthase gene requires both activator protein 1 and nuclear factor kappaB-binding sites. J Biol Chem, 1998. 273(35): p. 22201-8. 493. Aktan, F., iNOS-mediated nitric oxide production and its regulation. Life Sci, 2004. 75(6): p. 639-53. 494. Perkins, N.D., Integrating cell-signalling pathways with NF-kappaB and IKK function. Nat Rev Mol Cell Biol, 2007. 8(1): p. 49-62. 495. Yamamoto, M., et al., Essential role for TIRAP in activation of the signalling cascade shared by TLR2 and TLR4. Nature, 2002. 420(6913): p. 324-9. 496. Noman, A.S., et al., Thalidomide inhibits lipopolysaccharide-induced nitric oxide production and prevents lipopolysaccharide-mediated lethality in mice. FEMS Immunol Med Microbiol, 2009. 56(3): p. 204-11. 497. Yamamoto, M., et al., Role of adaptor TRIF in the MyD88-independent toll- like receptor signaling pathway. Science, 2003. 301(5633): p. 640-3.

193

498. Brieger, A., L. Rink, and H. Haase, Differential regulation of TLR-dependent MyD88 and TRIF signaling pathways by free zinc ions. J Immunol, 2013. 191(4): p. 1808-17. 499. Ostuni, R., I. Zanoni, and F. Granucci, Deciphering the complexity of Toll- like receptor signaling. Cell Mol Life Sci, 2010. 67(24): p. 4109-34. 500. Jacobs, A.T. and L.J. Ignarro, Cell density-enhanced expression of inducible nitric oxide synthase in murine macrophages mediated by interferon-beta. Nitric Oxide, 2003. 8(4): p. 222-30. 501. Adams, G.B. and D.T. Scadden, The hematopoietic stem cell in its place. Nat Immunol, 2006. 7(4): p. 333-7. 502. Morrison, S.J. and A.C. Spradling, Stem cells and niches: mechanisms that promote stem cell maintenance throughout life. Cell, 2008. 132(4): p. 598- 611. 503. Matsumoto, A., et al., p57 is required for quiescence and maintenance of adult hematopoietic stem cells. Cell Stem Cell, 2011. 9(3): p. 262-71. 504. Gurumurthy, S., et al., The Lkb1 metabolic sensor maintains haematopoietic stem cell survival. Nature, 2010. 468(7324): p. 659-63. 505. Taoudi, S., et al., Extensive hematopoietic stem cell generation in the AGM region via maturation of VE-cadherin+CD45+ pre-definitive HSCs. Cell Stem Cell, 2008. 3(1): p. 99-108. 506. Jing, D., et al., Hematopoietic stem cells in co-culture with mesenchymal stromal cells--modeling the niche compartments in vitro. Haematologica, 2010. 95(4): p. 542-50. 507. Kurtova, A.V., et al., Diverse marrow stromal cells protect CLL cells from spontaneous and drug-induced apoptosis: development of a reliable and reproducible system to assess stromal cell adhesion-mediated drug resistance. Blood, 2009. 114(20): p. 4441-50. 508. Jin, L., et al., CXCR4 up-regulation by imatinib induces chronic myelogenous leukemia (CML) cell migration to bone marrow stroma and promotes survival of quiescent CML cells. Mol Cancer Ther, 2008. 7(1): p. 48-58. 509. Roodhart, J.M., et al., Mesenchymal stem cells induce resistance to chemotherapy through the release of platinum-induced fatty acids. Cancer Cell, 2011. 20(3): p. 370-83. 510. Ramasamy, R., et al., Mesenchymal stem cells inhibit proliferation and apoptosis of tumor cells: impact on in vivo tumor growth. Leukemia, 2007. 21(2): p. 304-10. 511. Maestroni, G.J., E. Hertens, and P. Galli, Factor(s) from nonmacrophage bone marrow stromal cells inhibit Lewis lung carcinoma and B16 melanoma growth in mice. Cell Mol Life Sci, 1999. 55(4): p. 663-7. 512. Nakamura, K., et al., Antitumor effect of genetically engineered mesenchymal stem cells in a rat glioma model. Gene Ther, 2004. 11(14): p. 1155-64. 513. Cousin, B., et al., Adult stromal cells derived from human adipose tissue provoke pancreatic cancer cell death both in vitro and in vivo. PLoS One, 2009. 4(7): p. e6278. 514. Popovic, P.J., H.J. Zeh, 3rd, and J.B. Ochoa, Arginine and immunity. J Nutr, 2007. 137(6 Suppl 2): p. 1681S-1686S.

194

515. Angele, M.K., et al., L-arginine: a unique amino acid for restoring the depressed macrophage functions after trauma-hemorrhage. J Trauma, 1999. 46(1): p. 34-41. 516. Zea, A.H., et al., Decreased expression of CD3zeta and nuclear transcription factor kappa B in patients with pulmonary tuberculosis: potential mechanisms and reversibility with treatment. J Infect Dis, 2006. 194(10): p. 1385-93. 517. Kropf, P., et al., Arginase activity mediates reversible T cell hyporesponsiveness in human pregnancy. Eur J Immunol, 2007. 37(4): p. 935-45. 518. Rodriguez, P.C., et al., Regulation of T cell receptor CD3zeta chain expression by L-arginine. J Biol Chem, 2002. 277(24): p. 21123-9. 519. Gong, H., et al., Arginine deiminase inhibits cell proliferation by arresting cell cycle and inducing apoptosis. Biochem Biophys Res Commun, 1999. 261(1): p. 10-4. 520. Michurina, T., et al., Nitric oxide is a regulator of hematopoietic stem cell activity. Mol Ther, 2004. 10(2): p. 241-8. 521. Cheng, T., et al., Hematopoietic stem cell quiescence maintained by p21cip1/waf1. Science, 2000. 287(5459): p. 1804-8. 522. Alakel, N., et al., Direct contact with mesenchymal stromal cells affects migratory behavior and gene expression profile of CD133+ hematopoietic stem cells during ex vivo expansion. Exp Hematol, 2009. 37(4): p. 504-13. 523. Meads, M.B., L.A. Hazlehurst, and W.S. Dalton, The bone marrow microenvironment as a tumor sanctuary and contributor to drug resistance. Clin Cancer Res, 2008. 14(9): p. 2519-26. 524. Zimmerlin, L., et al., Mesenchymal stem cell secretome and regenerative therapy after cancer. Biochimie, 2013. 95(12): p. 2235-45. 525. Wei, Z., et al., Bone marrow mesenchymal stem cells from leukemia patients inhibit growth and apoptosis in serum-deprived K562 cells. J Exp Clin Cancer Res, 2009. 28: p. 141. 526. Konopleva, M., et al., Stromal cells prevent apoptosis of AML cells by up- regulation of anti-apoptotic proteins. Leukemia, 2002. 16(9): p. 1713-24. 527. Lin, Y.M., et al., Study on the bone marrow mesenchymal stem cells induced drug resistance in the U937 cells and its mechanism. Chin Med J (Engl), 2006. 119(11): p. 905-10. 528. Balakrishnan, K., et al., Influence of bone marrow stromal microenvironment on forodesine-induced responses in CLL primary cells. Blood, 2010. 116(7): p. 1083-91. 529. Du, J., et al., IFN-gamma-primed human bone marrow mesenchymal stem cells induce tumor cell apoptosis in vitro via tumor necrosis factor-related apoptosis-inducing ligand. Int J Biochem Cell Biol, 2012. 44(8): p. 1305-14. 530. Sage, E.K., et al., Systemic but not topical TRAIL-expressing mesenchymal stem cells reduce tumour growth in malignant mesothelioma. Thorax, 2014. 69(7): p. 638-47. 531. Simpson, E. and D. Roopenian, Minor histocompatibility antigens. Curr Opin Immunol, 1997. 9(5): p. 655-61. 532. Simpson, E., D. Scott, and P. Chandler, The male-specific histocompatibility antigen, H-Y: a history of transplantation, immune response genes, sex

195

determination and expression cloning. Annu Rev Immunol, 1997. 15: p. 39- 61. 533. Korngold, R. and J. Sprent, Lethal graft-versus-host disease after bone marrow transplantation across minor histocompatibility barriers in mice. Prevention by removing mature T cells from marrow. J Exp Med, 1978. 148(6): p. 1687-98. 534. Millrain, M., et al., Examination of HY response: T cell expansion, immunodominance, and cross-priming revealed by HY tetramer analysis. J Immunol, 2001. 167(7): p. 3756-64. 535. Scott, D.M., et al., Identification of a mouse male-specific transplantation antigen, H-Y. Nature, 1995. 376(6542): p. 695-8. 536. Fierz, W., et al., Non-H-2 and H-2-linked immune response genes control the cytotoxic T-cell response to H-Y. Immunogenetics, 1982. 15(3): p. 261-70. 537. Gallily, R., A. Warwick, and F.B. Bang, Effect of Cortisone of Genetic Resistance to Mouse Hepatitis Virus in Vivo and in Vitro. Proc Natl Acad Sci U S A, 1964. 51: p. 1158-64. 538. Li, Y.M., et al., Glycation products in aged thioglycollate medium enhance the elicitation of peritoneal macrophages. J Immunol Methods, 1997. 201(2): p. 183-8. 539. Neeper, M., et al., Cloning and expression of a cell surface receptor for advanced glycosylation end products of proteins. J Biol Chem, 1992. 267(21): p. 14998-5004. 540. Topley, N., et al., Activation of inflammation and leukocyte recruitment into the peritoneal cavity. Kidney Int Suppl, 1996. 56: p. S17-21. 541. Bellingan, G.J., et al., In vivo fate of the inflammatory macrophage during the resolution of inflammation: inflammatory macrophages do not die locally, but emigrate to the draining lymph nodes. J Immunol, 1996. 157(6): p. 2577-85. 542. Rajakariar, R., et al., Novel biphasic role for lymphocytes revealed during resolving inflammation. Blood, 2008. 111(8): p. 4184-92. 543. Devuyst, O., P.J. Margetts, and N. Topley, The pathophysiology of the peritoneal membrane. J Am Soc Nephrol, 2010. 21(7): p. 1077-85. 544. Abbitt, K.B., et al., Antibody ligation of murine Ly-6G induces neutropenia, blood flow cessation, and death via complement-dependent and independent mechanisms. J Leukoc Biol, 2009. 85(1): p. 55-63. 545. Fleming, T.J., M.L. Fleming, and T.R. Malek, Selective expression of Ly-6G on myeloid lineage cells in mouse bone marrow. RB6-8C5 mAb to granulocyte- differentiation antigen (Gr-1) detects members of the Ly-6 family. J Immunol, 1993. 151(5): p. 2399-408. 546. Goldrath, A.W., L.Y. Bogatzki, and M.J. Bevan, Naive T cells transiently acquire a memory-like phenotype during homeostasis-driven proliferation. J Exp Med, 2000. 192(4): p. 557-64. 547. Rose, S., A. Misharin, and H. Perlman, A novel Ly6C/Ly6G-based strategy to analyze the mouse splenic myeloid compartment. Cytometry A, 2012. 81(4): p. 343-50. 548. Geissmann, F., S. Jung, and D.R. Littman, Blood monocytes consist of two principal subsets with distinct migratory properties. Immunity, 2003. 19(1): p. 71-82.

196

549. Austyn, J.M. and S. Gordon, F4/80, a monoclonal antibody directed specifically against the mouse macrophage. Eur J Immunol, 1981. 11(10): p. 805-15. 550. Gonzalez-Rey, E., et al., Human adult stem cells derived from adipose tissue protect against experimental colitis and sepsis. Gut, 2009. 58(7): p. 929-39. 551. Augello, A., et al., Cell therapy using allogeneic bone marrow mesenchymal stem cells prevents tissue damage in collagen-induced arthritis. Arthritis Rheum, 2007. 56(4): p. 1175-86. 552. Gonzalez-Rey, E., et al., Human adipose-derived mesenchymal stem cells reduce inflammatory and T cell responses and induce regulatory T cells in vitro in rheumatoid arthritis. Ann Rheum Dis, 2010. 69(1): p. 241-8. 553. Ohnishi, S., et al., Transplantation of mesenchymal stem cells attenuates myocardial injury and dysfunction in a rat model of acute myocarditis. J Mol Cell Cardiol, 2007. 42(1): p. 88-97. 554. Le Blanc, K., et al., Treatment of severe acute graft-versus-host disease with third party haploidentical mesenchymal stem cells. Lancet, 2004. 363(9419): p. 1439-41. 555. Ringden, O., et al., Mesenchymal stem cells for treatment of therapy- resistant graft-versus-host disease. Transplantation, 2006. 81(10): p. 1390-7. 556. Sung, H.C., et al., Cognate antigen stimulation generates potent CD8(+) inflammatory effector T cells. Front Immunol, 2013. 4: p. 452. 557. Lam, D., D. Harris, and Z. Qin, Inflammatory mediator profiling reveals immune properties of chemotactic gradients and macrophage mediator production inhibition during thioglycollate elicited peritoneal inflammation. Mediators Inflamm, 2013. 2013: p. 931562. 558. McCafferty, D.M., et al., Inducible nitric oxide synthase plays a critical role in resolving intestinal inflammation. Gastroenterology, 1997. 112(3): p. 1022-7. 559. Cho, K.S., et al., Adipose-derived stem cells ameliorate allergic airway inflammation by inducing regulatory T cells in a mouse model of asthma. Mediators Inflamm, 2014. 2014: p. 436476. 560. Ruster, B., et al., Mesenchymal stem cells display coordinated rolling and adhesion behavior on endothelial cells. Blood, 2006. 108(12): p. 3938-44. 561. Ries, C., et al., MMP-2, MT1-MMP, and TIMP-2 are essential for the invasive capacity of human mesenchymal stem cells: differential regulation by inflammatory cytokines. Blood, 2007. 109(9): p. 4055-63. 562. Ohtaki, H., et al., Stem/progenitor cells from bone marrow decrease neuronal death in global ischemia by modulation of inflammatory/immune responses. Proc Natl Acad Sci U S A, 2008. 105(38): p. 14638-43. 563. Ortiz, L.A., et al., Interleukin 1 receptor antagonist mediates the antiinflammatory and antifibrotic effect of mesenchymal stem cells during lung injury. Proc Natl Acad Sci U S A, 2007. 104(26): p. 11002-7. 564. Parsonage, G., et al., A stromal address code defined by fibroblasts. Trends Immunol, 2005. 26(3): p. 150-6. 565. McGettrick, H.M., et al., Tissue stroma as a regulator of leukocyte recruitment in inflammation. J Leukoc Biol, 2012. 91(3): p. 385-400. 566. Munir, H., et al., Analyzing the Effects of Stromal Cells on the Recruitment of Leukocytes from Flow. J Vis Exp, 2015(95).

197

567. Luu, N.T., et al., Crosstalk between mesenchymal stem cells and endothelial cells leads to downregulation of cytokine-induced leukocyte recruitment. Stem Cells, 2013. 31(12): p. 2690-702. 568. Maggini, J., et al., Mouse bone marrow-derived mesenchymal stromal cells turn activated macrophages into a regulatory-like profile. PLoS One, 2010. 5(2): p. e9252. 569. Ravishankar, B., et al., Tolerance to apoptotic cells is regulated by indoleamine 2,3-dioxygenase. Proc Natl Acad Sci U S A, 2012. 109(10): p. 3909-14. 570. Gupta, N., et al., Intrapulmonary delivery of bone marrow-derived mesenchymal stem cells improves survival and attenuates endotoxin- induced acute lung injury in mice. J Immunol, 2007. 179(3): p. 1855-63. 571. Caplan, A.I. and J.E. Dennis, Mesenchymal stem cells as trophic mediators. J Cell Biochem, 2006. 98(5): p. 1076-84. 572. Duijvestein, M., et al., Autologous bone marrow-derived mesenchymal stromal cell treatment for refractory luminal Crohn's disease: results of a phase I study. Gut, 2010. 59(12): p. 1662-9. 573. Bueno, C., et al., Bone marrow mesenchymal stem cells from patients with aplastic anemia maintain functional and immune properties and do not contribute to the pathogenesis of the disease. Haematologica, 2014. 99(7): p. 1168-75. 574. Murphy, J.M., et al., Reduced chondrogenic and adipogenic activity of mesenchymal stem cells from patients with advanced osteoarthritis. Arthritis Rheum, 2002. 46(3): p. 704-13. 575. Scharstuhl, A., et al., Chondrogenic potential of human adult mesenchymal stem cells is independent of age or osteoarthritis etiology. Stem Cells, 2007. 25(12): p. 3244-51. 576. Kebriaei, P., et al., Adult human mesenchymal stem cells added to corticosteroid therapy for the treatment of acute graft-versus-host disease. Biol Blood Marrow Transplant, 2009. 15(7): p. 804-11. 577. Wu, J., et al., Intravenously administered bone marrow cells migrate to damaged brain tissue and improve neural function in ischemic rats. Cell Transplant, 2008. 16(10): p. 993-1005. 578. Suda, T., K. Takubo, and G.L. Semenza, Metabolic regulation of hematopoietic stem cells in the hypoxic niche. Cell Stem Cell, 2011. 9(4): p. 298-310. 579. Petit-Bertron, A.F., et al., H4 histamine receptors mediate cell cycle arrest in growth factor-induced murine and human hematopoietic progenitor cells. PLoS One, 2009. 4(8): p. e6504. 580. Dvorak, H.F., Tumors: wounds that do not heal. Similarities between tumor stroma generation and wound healing. N Engl J Med, 1986. 315(26): p. 1650-9. 581. Mantovani, A., et al., Cancer-related inflammation. Nature, 2008. 454(7203): p. 436-44.

198