Open Research Online The Open University’s repository of research publications and other research outputs

Expression chemokines and G protein-coupled receptors in myeloid cells

Thesis

How to cite:

Skelton, Lisa (2003). Expression chemokines and G protein-coupled receptors in myeloid cells. PhD thesis The Open University.

For guidance on citations see FAQs.

c 2003 Lisa Skelton

Version: Version of Record

Link(s) to article on publisher’s website: http://dx.doi.org/doi:10.21954/ou.ro.0000f566

Copyright and Moral Rights for the articles on this site are retained by the individual authors and/or other copyright owners. For more information on Open Research Online’s data policy on reuse of materials please consult the policies page. oro.open.ac.uk Expression of chemokines and G protein- coupled receptors in myeloid cells

Lisa Skelton BSc (Hons)

A thesis submitted in partial fulfillment of the requirements of the Open University for the degree of Doctor of Philosophy

October 2003

Sponsoring Establishment: Department of Cell Biology Celltech R&D Ltd. 216 Bath Road, Slough, Berkshire

Collaborating Establishment: Bone and Mineral Centre University College London 5 University Street, London ProQuest Number: C816612

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest

ProQ uest C816612

Published by ProQuest LLC (2019). Copyright of the Dissertation is held by the Author.

All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC.

ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 4 8 1 0 6 - 1346 Publication arising from this thesis:

Skelton, L., Cooper, M., Murphy, M., Platt, A.

Human immature monocyte-derived dendritic cells express the G protein- coupled receptor GPR105 (KIAAOOOl, P2Y14) and increase intracellular calcium in response to its agonist, uridine diphosphoglucose.

J. Immunol. 2003 Aug 15; 171(4): 1941-9. Abstract

The myeloid cell lineage represents a diverse collection of cell types arising from a common progenitor. Two extreme phenotypes of this lineage belong to the osteoclast and the dendritic cell.

This thesis evaluates differential expression of a discrete selection of genes in these two cell types, as they differentiate from their monocyte precursor, utilising well-characterized in vitro differentiation systems and microarray gene technology.

A custom made microarray, including all known chemokine and chemokine receptor genes and related G protein-coupled seven transmembrane (7TM) receptors and ligands, was employed.

Expression of transcripts encoding chemokine and chemokine receptors by dendritic cells, and a

study of protein expression / production by these cells, served as validation for the microarray gene technology used. In particular, the array data revealed two interesting findings, which were

investigated further.

GPR105 gene expression by immature DCs was confirmed by Northern blot analysis. UDP-glucose,

an agonist for GPR105, induced calcium responses uniquely in immature DCs, correlating with the

gene expression data. This response demonstrated partial sensitivity to pertussis toxin, implicating a

7TM receptor linked to a Gia containing G protein, such as GPR105. Moreover, immature DCs

from some donors treated with UDP-glucose exhibited an increase in expression of the co­

stimulatory molecule CD86, which correlated with the intensity of an UDP-glucose-induced

calcium flux, suggesting a role for this GPCR and its agonist in DC activation.

Production of the angiogenic chemokine CXCL5 by in v/fro-derived osteoclast cultures was

confirmed by ELISA and attributed to multinuclear cells by immunocytochemistry. Supernatants

from osteoclast cultures induced chemotaxis of neutrophils, which was decreased in the presence of

an antibody to CXCL5. A role for osteoclast produced CXCL5 in angiogenesis is discussed. Acknowledgements

I would like to thank my supervisors for their help and support throughout this project, all had very different roles to play. Firstly, Dr. Marianne Murphy for encouraging me from the outset, giving me the confidence to propose this project, helping to convince the powers that be to support me in my efforts, but mostly for sticking with me when maybe there were other more important things on her mind! Then my external supervisor Prof. Mike Horton for his pragmatic support throughout, some very pertinent advice which, incidentally, I wish I had taken more notice of and for always making me feel welcome in his lab. Lastly, Dr. Andrew McKnight for seeing me down the home straight, sorry the proof-reading clashed with the Champions League!

I would also like to thank those around me in the lab, for their help and encouragement (well

sometimes) along the way. The Cell Biology lab at Celltech is where most of my time was spent

and I thank Emma, Marianne, Ski, Nobby and Ken for the hours of banter that kept me sane (not so

sure about them though)! I also thank Dr. Adam Platt for his help with the microarray work and the

JI publication; Adam your enthusiasm and drive is really infectious, it definitely helped me get this

far. Then of course there is Bemie, it was great to have someone else in the same boat, and it

couldn’t have been anyone better.

Then comes Mum and Dad, thanks guys, you are the best and I promise to invite you round for

dinner more often now!

Finally, to my husband Mark, I didn’t think you would understand, but you did, your support was

amazing, I couldn’t have done this without you. Abbreviations

ADCC antibody-dependant cell-mediated cytotoxicity AEC 3-amino-9-ethylcarbazole ATP adenosine triphosphate bp base pair BDCA blood dendritic cell antigen BLC B lymphocyte chemoattractant BSA bovine serum albumin cAMP cyclic adenosine monophosphate CCD charge-coupled device CTACK cutaneous T cell attracting chemokine CTR calcitonin receptor dATP deoxyadenosine 5’-triphosphate DC dendritic cell dCTP deoxycytidine 5’-triphosphate dGTP deoxyguanosine 5'-triphosphate DMSO dimethyl sulphoxide DNA deoxyribonucleic acid DTT dithiothreitol dTTP deoxythymidine 5’-triphosphate EDG endothelial differentiation gene EDTA ethylenediamine tetra-acetic acid EGF epidermal growth factor EEC Epstein-Barr-virus-induced gene 1 ligand chemokine ELISA enzyme linked immunosorbent assay EMRl EGF module-containing mucin-like hormone receptor 1 ENA-78 epithelial-derived neutrophil activating protein 78 EPO erythropoietin ESTs expressed sequence tags FACS fluorescence-activated cell sorter FCR Fc receptor FCS foetal calf serum FITC fluorescein isothiocyanate FIU fluorescence intensity units FLIPR fluorometric imaging plate reader FSC / SSC forward scatter / side scatter GCP-2 granulocyte chemotactic protein 2 GM-CSF granulocyte macrophage-colony stimulating factor GPCR G protein-coupled receptor G3PDH glyceraldehyde-3 -phosphate dehydrogenase GRO- growth-related oncogene GTP guanosine triphosphate HA haemagglutinin HCC- haemofiltrate CC chemokine- HEK293 human embryonic kidney 293 HIV human immunodeficiency virus HLA human leukocyte antigen HMVEC human microvascular endothelial cells HUVEC human umbilical vein endothelial cells HRP horse radish peroxidase if n -y interferon-gamma IL- interleukin IPIO IFN-gamma-Inducible Protein-10 LEK lipopolysaccharide-induced CXC chemokine LPA lysophosphatidic acid EPS lipopolysaccharide mAb monoclonal antibody MCM monocyte-conditioned medium MCP- monocyte chemoattractant protein- M-CSF macrophage-colony stimulating factor MDC macrophage-derived chemokine MEG mucosae associated epithelial chemokine MFI mean fluorescence intensity MHC major histocompatability complex MIG monokine induced by IFN-gamma MIP- macrophage inhibitory protein- MLR mixed leukocyte reaction MNC multinucleated cells MPIF- myeloid progenitor inhibitory factor- mRNA messenger ribonucleic acid NAP-2 neutrophil activating peptide 2 OPG osteoprotegrin P&S penicillin and streptomycin PARC pulmonary and activation-regulated cytokine PAR- protease-activated receptor- PBM peripheral blood monocytes PBMC peripheral blood mononuclear cells PBS phosphate buffered saline PCR polymerase chain reaction PE phycoerythrin PF4 platelet factor 4

PGE2 prostaglandin E 2

RANK receptor activator of NF-k B

RANKE receptor activator of NF-k B ligand RANTES regulated upon activation normal T cell expressed and secreted rRNA ribosomal RNA RNA ribonucleic acid RT- reverse transcription- SAC Staphylococcus aureus Cowan’s strain SAGE serial analysis of gene expression SCF stem cell factor SD standard deviation SDF-1 stromal-cell-derived factor-1 SEM standard error of mean SDS sodium dodecyl sulphate SIP sphingosine 1-phosphate SEC secondary lymphoid tissue chemokine SSC sodium chloride/ sodium citrate buffer TARC thymus-associated chemokine TECK thymus-expressed chemokine TGF-P transforming growth factor-beta Thl T helper 1 Th2 T helper 2 TM transmembrane TMB tetramethylbenzidine TNF-a tumour necrosis factor-alpha TRAP tartrate resistant acid phosphatase UDF uridine diphosphate UDP-glucose uridine diphosphoglucose UTP uridine triphosphate UTR untranslated region UV ultraviolet VEGF vascular endothelial growth factor 7TM 7 transmembrane Table of contents

ABSTRACT...... 3

1 CHAPTER 1 INTRODUCTION...... 14

1.1 Haematopoiesis, bone and the immune system______15 1.2 Myeloid cell differentiation pathways: macrophages, dendritic cells and osteoclasts 19 1.2.1 Macrophages ...... 19 1.2.2 Dendritic cells ...... 20 1.2.3 Osteoclasts ...... 23 1.2.4 Macrophages, dendritic cells and osteoclasts ...... 25 1.3 DC isolation and differentiation pathways______26 1.4 Osteoclast isolation and differentiation pathways______30 1.5 Gene expression: mRNA transcript analysis______33 1.6 mRNA transcript analysis of a selection of genes in myeloid cells______35 1.7 The GPCR superfamily ______35 1.8 Chemokine receptors______37 1.8.1 CXC receptors ...... 37 1.8.2 CC receptors ...... 38 1.9 Chemokines______39 1.10 Chemokines and their receptors in DC biology______43 1.11 Chemokines and their receptors in osteoclast biology ______44 1.12 Other GPCRs and ligands: an overview______45 1.13 Project Aim s______47 2 CHAPTER 2 MATERIALS AND METHODS...... 49

2.1 Materials______50 2.1.1 General reagents ...... 50 2.2 Methods______50 2.2.1 Purification of human PBM C ...... 50 2.2.2 Purification of human PBM ...... 50 2.2.3 Monocyte-conditioned medium ...... 51 2.2.4 DC generation ...... 52 2.2.5 Flow cytometry...... 52 2.2.6 Endocytosis assays ...... 53 2.2.7 Allogeneic T cell proliferation and IFN-y production ...... 53 2.2.8 IL-12 production by DCs ...... 54 2.2.9 Osteoclast generation ...... 54 2.2.10 Osteoclast culture staining ...... 55 2.2.11 Osteoclast pit formation ...... 56 2.2.12 RNA extraction and characterisation ...... 57 2.2.13 RT-PCR for RANK and calcitonin receptor ...... 58 2.2.14 cDNA probe synthesis, microarray hybridisation and quantification ...... 60 2.2.15 Northern blot hybridization ...... 61 2.2.16 DC chemokine ELIS As ...... 61 2.2.17 Characterisation of HEK293 cell lines ...... 62 2.2.18 Ca^^ flux measurement ...... 62 2.2.19 Preparation of necrotic cell supernatant ...... 63 2.2.20 Osteoclast chemokine ELIS As ...... 64 2.2.21 Endothelial cell chemotaxis ...... 64 2.2.22 Neutrophil chemotaxis ...... 65 2.2.23 Immunocytochemistry ...... 65 3 CHAPTER 3 IN KTTRD-DERIVED OSTEOCLASTS AND DENDRITIC CELLS. 66

3.1 Introduction ______67 3.2 Results______69 3.2.1 Purification of monocytes from peripheral blood ...... 69 3.2.2 Differentiation and characterisation of dendritic cells from peripheral blood monocytes ...... 71 3.2.3 Differentiation and characterisation of osteoclasts from peripheral blood monocytes 77 3.3 Discussion ______83 4 CHAPTER 4 MICRO ARRAY PROFILING OF MONOCYTES, OSTEOCLASTS AND DENDRITIC CELLS...... 86

4.1 Introduction ______87 4.2 Results______90 4.2.1 Characterisation of monocyte, DC and osteoclast samples for RNA extraction ...... 90 4.2.2 Gene array profiling of dendritic cell, osteoclast and monocyte populations ...... 94 4.2.3 Differentially expressed genes ...... 96 4.2.4 Correlation of gene array and protein expression data in dendritic cells ...... 98 4.2.5 Chemokine and chemokine receptor expression in osteoclasts ...... 104 4.2.6 7TM GPCRs expressed in monocytes, osteoclasts and dendritic cells ...... 106 4.3 Discussion ______110 5 CHAPTER 5 GPR105 IN DENDRITIC C E LLS...... 122

5.1 Introduction ______123 5.2 Results______125 5.2.1 GPRl 05 transcript expression in DCs ...... 125 5.2.2 UDP-glucose-induced calcium signalling in a GPRl05-transfected cell line ...... 127 5.2.3 Correlation of UDP-glucose-induced intracellular calcium release with GPRl 05 expression ...... 131 5.2.4 Phenotypic changes in immature DCs in response to UDP-glucose ...... 133 5.2.5 The effect of endotoxin contamination on UDP-glucose-induced CD86 expression 138 5.2.6 Endotoxin free UDP-glucose ...... 141 5.2.7 The effect of pertussis toxin pre-treatment on endotoxin-free UDP-glucose-induced calcium flux in immature DC ...... 146 5.2.8 Intracellular calcium release from GPRl 05-expressing HEK293 cells in response to necrotic cell components ...... 148 5.3 Discussion ______150 6 CHAPTER 6 CXCL5 IN OSTEOCLASTS...... 157

10 6.1 Introduction ______158 6.2 Results ;______161 6.2.1 Transcription and translation of CXC chemokine genes in osteoclast cultures 161 6.2.2 Immunocytochemical analysis of osteoclast chemokine production ...... 164 6.2.3 Chemotaxis to osteoclast supernatants ...... 166 6.2.4 Induction of CXCL5 and CXCL8 production by angiogenic and inflammatory mediators ...... 169 6.3 Discussion ______171 CHAPTER 7 GENERAL DISCUSSION ...... 175

7.1 Overview______176 7.2 Discussion ______180 7.3 Summary ______187 CHAPTER 8 REFERENCE LIST 196

11 Table of Figures, Tables and Appendiees

Figure 2.1 Haematopoiesis ...... 18 Figure 2.2 Three subsets of human dendritic cells ...... 22 Figure 2.3 A schematic representation of osteoclast molecules and functions ...... 24 Figure 2.4 The differentiation and maturation of in vitro monocyte-derived DCs ...... 29 Figure 2.5 Schematic for the in vitro differentiation of osteoclasts and characteristics of these cells ...... 32 Figure 3.1 Two colour flow cytometric analysis of PBMC and purified monocytes ...... 70 Figure 3.2 Cell surface phenotypes of immature and mature D Cs ...... 73 Figure 3.3 Comparison of antigen uptake capabilities of immature and mature DCs ...... 74 Figure 3.4 Proliferation of naïve T cells in response to allogeneic immature or mature DCs .....74 Figure 3.5 IFN-y production from T cells in response to allogeneic immature or mature DCs...75 Figure 3.6 IL-12 p40 production from DCs following maturation with monocyte-conditioned medium...... 75 Figure 3.7 Induction of IL-12 p70 from monocyte-derived DCs ...... 76 Figure 3.8 Staining of M-CSF-plus RANKL-derived osteoclast preparations for aV(33-positive and TRAP-positive MNC ...... 78 Figure 3.9 TRAP-positive multinucleated cell formation in osteoclast differentiation cultures..79 Figure 3.10 TRAP-positive multinuclear cells and resorption lacunae on dentine slices ...... 80 Figure 3.11 Time course for formation of TRAP-positive multinuclear cells and resorption lacunae on dentine slices ...... 81 Figure 3.12 RT-PCR analysis of CTR, G3PDH and RANK transcripts mRNA expression in monocytes and osteoclasts ...... 82 Figure 4.1 Schematic of a differential gene expression hybridisation experiment ...... 89 Figure 4.2 Characterisation of dendritic cell preparations for gene array profiling ...... 91 Figure 4.3 Characterisation of osteoclast preparations for gene array profiling ...... 92 Figure 4.4 The integrity of RNA preparations subjected to microarray profiling, as examined by formaldehyde agarose gel electrophoresis ...... 93 Figure 4.5 Hierarchical clustering of mRNA transcript expression data acquired by gene microarray profiling ...... 95 Figure 4.6 Chemokine receptor mRNA and cell surface expression levels ...... 100 Figure 4.7 CCL13, CCL15 and CCL23 mRNA and protein levels in DC populations ...... 101 Figure 4.8 CCL3, CCL4 and SCYA3L mRNA and protein levels in DC populations ...... 102 Figure 4.9 CCL17, CCL22 and CCL18 mRNA and protein levels in DC populations ...... 103 Figure 4.10 CCL2 mRNA and protein levels in DC populations ...... 104 Figure 4.11 CCL3, SCYA3L, CCL20, CCL4, CXCL8 and CCR2 gene expression in monocytes and osteoclasts ...... 105 Figure 4.12 CXCL5 gene expression in monocytes, osteoclasts and DC ...... 106 Figure 4.13 Differential GPCR gene expression in monocytes and osteoclasts ...... 107 Figure 4.14 Differential GPCR gene expression in DC populations ...... 108 Figure 4.15 GPRl 05 mRNA expression in myeloid cells ...... 109 Figure 5.1 GPRl 05 transcript expression in DC populations ...... 125 Figure 5.2 Northern blot analysis of GPRl 05 and (3-actin mRNA expression ...... 126 Figure 5.3 Expression of HA-tagged GPRl 05 in HEK293 transfectants ...... 128 Figure 5.4 Intracellular calcium release in a GPR105/Gai6 expressing and control HEK293 cell lines ...... 129 Figure 5.5 Intracellular calcium release in response to UDP-glucose in monocytes and DCs ..132 Figure 5.6 CD80, CD86, HLA-DR and CD83 expression on UDP-glucose-treated DCs ...... 135 Figure 5.7 The effect of UDP-glucose on transcript expression profiles in DCs ...... 136

12 Figure 5.8 The effect of UDP-glucose on the endocytic ability of DCs ...... 137 Figure 5.9 The effect of polymyxin B on UDP-glucose-induced CD86 expression by DCs.... 139 Figure 5.10 The effect of polymyxin B on histamine, CCL3 and UDP-glucose-initiated intracellular calcium release in immature DCs ...... 140 Figure 5.11 The effect of endotoxin free UDP-glucose on CD86 expression by immature DCs 142 Figure 5.12 The effect of UDP-glucose and LPS on CD86 expression and CD83 expression and FITC-dextran uptake by DCs ...... 143 Figure 5.13 The effect of UDP-glucose pre-treatment of DCs, with or without LPS, on allogeneic T cell proliferation and IFN-y production ...... 144 Figure 5.14 Correlation of UDP-glucose-induced CD86 expression and UDP-glucose-induced calcium flux...... 145 Figure 5.15 The effect of pertussis toxin pre-treatment on the calcium responses of immature DCs to histamine, CCL3 and endotoxin free UDP-glucose ...... 147 Figure 5.16 GPRl 05-expressing HEK293 and control HEK293 responses to necrotic cell supernatants ...... 149 Figure 6.1 Expression of CXCLl, CXCL5, CXCL7, CXCL8, CXCL9 and CXCLIO mRNA transcripts in monocytes and osteoclasts ...... 162 Figure 6.2 Levels of CXC chemokines in the supernatants of osteoclast cultures with or without RANKL...... 163 Figure 6.3 CXCL5 and CXCL8 production in osteoclast cultures ...... 165 Figure 6.4 Endothelial cell chemotaxis to 10% serum ...... 167 Figure 6.5 Chemotaxis of human peripheral blood neutrophils in response to CXCL5 and CXCL8 ...... 167 Figure 6.6 Blocking of neutrophil chemotaxis to osteoclast supernatants by anti-CXCL5 and anti- CXCL8 antibodies ...... 168 Figure 6.7 The effect of VEGF on CXCL5 and CXCL8 production from osteoclast cultures.. 169 Figure 6.8 The induction of CXCL5 and CXCL8 from osteoclast cultures by inflammatory mediators ...... 170

Table 1.1 Human chemokines, their receptors and main functions ...... 42

Appendix 1 Summary tables of microarray gene expression data ...... 188 Appendix 2 Genes represented on the microarray with GenBank accession numbers ... 194

13 Chapter 1

Introduction

14 Introduction

1.1 Haematopoiesis, bone and the immune system

Blood cells and certain other cells located throughout the body are continuously regenerated throughout life by a process called haematopoiesis. Many haematopoietic cells are short lived, some surviving for only a day or two, and thus haematopoiesis serves to maintain a steady renewal of these cells on physiological demand. Reconstitution of the haematopoietic system is believed to be the function of a single haematopoietic precursor cell called the pluripotent stem cell (Morrison et al., 1995). In addition to being self-renewing, haematopoietic stem cells undergo multi-lineage differentiation driven by cytokines and cell-cell interactions. There is a vast literature regarding this differentiation process. Figure 1.1 summarises some of the relevant steps in the process.

The pluripotent haematopoietic stem cell undergoes lineage commitment and subsequent sequential differentiation along distinct lineage pathways, giving rise to all blood cell types. The common lymphoid progenitor cells differentiate into NK cells, B lymphocytes and T lymphocytes. The stem cells committed to the myeloid lineage (common myeloid progenitors) differentiate into platelets, red blood cells, granulocytes, monocytes, macrophages, and myeloid dendritic cells (DCs).

Osteoclasts and other specialised cell types such as microglial cells, Kupffer cells and alveolar macrophages also develop from the myelomonocytic lineage (Kaur et al., 2001; Nicholson et al.,

2000).

In general, haematopoietic cells are highly mobile, migrating within the cardiovascular system or the lymphatic system. Many of these cells also cross-migrate between these two circulatory systems and many migrate directly into peripheral tissues, particularly under the influence of inflammatory cytokines and chemokines. These cells represent the immune system, constituted by a surprising variety of different cell types disseminated throughout the body and collectively defining our

15 capacity to mount an immune response to infectious pathogens such as viruses and bacteria. The

immune system is typically divided into two categories, innate and adaptive, although these

distinctions are not mutually exclusive. Innate immunity refers to non-specific defence mechanisms that come into play immediately or within hours of an antigen's appearance in the body. These mechanisms include physical barriers such as skin, chemicals in the blood, activation of the complement system and release of cytokines and chemokines. The cells involved in the innate immune system include: cells that release inflammatory mediators, such as basophils, mast cells and eosinophils; NK cells that specialise in killing tumour cells and virally infected cells; and phagocytic cells, such as neutrophils, monocytes and macrophages, which bind to microorganisms using primitive non-specific recognition systems, internalise them and destroy them. Adaptive immunity refers to an antigen-specific immune response. The adaptive immune response is more complex than the innate. The antigen first must be processed and recognised. Once an antigen has been recognised, the adaptive immune system creates an army of immune cells specifically designed to attack that antigen. Adaptive immunity also includes immunological memory that makes future responses against a specific antigen more efficient and involves: antigen-presenting cells such as macrophages and dendritic cells; the activation and proliferation of antigen-specific B- lymphocytes; the activation and proliferation of antigen-specific T lymphocytes; the production of antibody molecules, cytotoxic T lymphocytes (CTLs), activated macrophages, NK cells and cytokines.

The interdependence of bone and the haematopoietic system has long been apparent. Bone remodelling depends upon the resorption of bone by specialised osteoclasts, which arise from the haematopoietic myelomonocytic lineage. Indeed development of bone marrow, the home of haematopoiesis in the adult, depends upon the excavation of cartilage and bone by osteoclasts.

Furthermore, mice lacking c-Fos, a component of the AP-1 transcription factor complex, demonstrate an osteopetrotic phenotype and altered haematopoiesis, with decreased numbers of

16 splénocytes and thymocytes (Wang et al., 1992). There is also evidence that megakaryocytes play a role in the regulation of bone remodelling, particularly as they express the osteoclastogenic factor

RANKL (Receptor activator of NF-k B ligand) (Kartsogiannis et al., 1999).

Relationships between bone remodelling and the immune system are perhaps less obvious.

Osteopetrotic syndromes, arising from a deficiency of bone resorption, often include associated immune defects. However, this is generally a secondary consequence of the failure to establish a normal haematopoietic environment. Glucocorticoids and other immunosuppressive drugs are known to cause a net loss of bone that can predispose to fracture (Canalis, 2003). Similarly, many cytokines and other regulatory molecules that are important regulators of immune cell differentiation and maturation also have potent effects on osteoblasts and osteoclasts. However, it is for the most part unclear whether the effects on bone remodelling reflect interactions between bone cells and immune cells or, instead, independent effects of the agents on the two systems. Perhaps the strongest link of bone to the immune system has emerged from the discovery of the RANK

(Receptor activator of NF-k B) / RANKL / OPG (osteoprotegrin) system. RANKL has been identified as a key osteoclast differentiation factor, a regulator of dendritic cell functions and is essential for lymph node organogenesis (Kong et al., 1999a). T cells express RANKL and can directly trigger osteoclastogenesis and bone loss (Kong et al., 1999b). The contribution of bone marrow T cells relative to stromal cells in osteoclastogenesis is unknown. Intriguing also is the common progenitor of two highly specialised cell types with central roles in bone and the immune system, the osteoclast and the dendritic cell respectively.

17 Adapted from; Kuby Immunology, 4*^ edition, Chapter 3: Antigens

To cell

Thymocyte T progenitor

Tr cell

B progenitor Lymphoid B cell stem cell

Dendritic cell

renewal

Monocyte

Granulocyte MacrophagePluripotent Stem cell monocyte progenitor

Granulocyte

Osteoclast Eosinophil progenitor

Eosinophil

Basophil progenitor

Myeloid stem cell Basophil

Platelets Megakaryoc

Erythrocyte

Erythroid progenitor

Figure 1.1 Haematopoiesis Haematopoietic cells are continuously regenerated throughout life by the process of haematopoiesis: steadily renewing these cells on physiological demand. Haematopoiesis is believed to arise from the function of a single precursor cell called the pluripotent stem cell. In addition to being self-renewing, stem cells undergo multi-lineage differentiation. Stem cells are shown in grey; progenitor cells in green and fully differentiated cell types in blue.

18 1.2 Myeloid cell differentiation pathways: macrophages, dendritic

cells and osteoclasts

There are three major pathways of myeloid cell differentiation; erythroid, megakaryocytic and phagocytic. The erythroid lineage gives rise to red blood cells; the megakaryocytic lineage generates blood platelets, while the phagocytic pathway undergoes complex subdivisions, firstly into the monocytic and the granulocytic lineages. The granulocyte lineage divides into the neutrophilic, eosinophilic, and basophilic pathways. Blood monocytes, which mature in all tissues and organs of the body to various types of macrophages, are also capable of differentiating into dendritic cells and osteoclasts (Fig. 1.1).

1.2.1 Macrophages

Macrophages can be classified into either resident or inflammatory macrophages. Resident macrophages, such as those found in the connective tissue (histiocytes), liver (Kupffer cells), lung

(alveolar macrophages), kidney (mesangial cells) and peritoneal macrophages to name but a few are renewed by local proliferation and differentiation of progenitor cells. Inflammatory macrophages are characterized by various specific markers, including peroxidase activity, and are derived exclusively from recruited monocytes. Macrophages are a population of ubiquitously distributed mononuclear phagocytes responsible for numerous homeostatic, immunological and inflammatory processes. Their range of tissue distribution makes these cells perfectly suited to provide an immediate defense against foreign pathogens prior to adaptive immune responses. Since macrophages participate in both specific immunity via antigen processing and presentation and innate immunity against bacterial, viral and fungal pathogens, it is not surprising that macrophages display a wide range of functional and phenotypic properties. The microbiocidal and cytotoxic activity of macrophages occurs mainly via the production of reactive oxygen intermediates,

allowing the killing of phagocytosed microorganisms. In addition, macrophages are capable of

19 antibody-dependent cell-mediated cytotoxicity (ADCC), insofar as they are able to kill or damage antibody-opsonised extracellular microbes. Macrophages may also take part in the initiation of specific T cell activation through the processing and presentation of foreign antigen. Finally these cells are central effector and regulatory cells of the inflammatory response, acting through the release of numerous immunoregulatory or inflammatory cytokines (Gordon, 1996a; Gordon,

1996b).

1.2.2 Dendritic cells

The term dendritic cell represents a heterogeneous group of cells, which, under different microenvironmental conditions, can induce such contrasting states as T cell driven immunity and T cell tolerance. That is, the immune system is uniquely characterised by its highly diverse and clonally expressed repertoire of lymphocyte receptors that can recognise a very broad spectrum of both foreign and self-antigens. Those cells with receptors for self-antigens are eliminated, or tolerised, in healthy individuals by a variety of active mechanisms to prevent pathogenic autoimmunity. These same mechanisms operate during the induction of tolerance to foreign antigens and may be exploited to inactivate specific, detrimental immune responses without compromising the ability to respond to the remaining universe of pathogens, because tolerance is induced only in those lymphocytes that recognise the offending antigens. Therefore, presentation of processed antigen to T cells by dendritic cells can result in either tolerance induction or initiation of an immune response.

Despite the heterogeneity of human dendritic cells they have many factors in common: they are all able to internalise and process antigen; they possess the ability to migrate to lymphoid tissue in response to danger signals and they have the ability to stimulate a primary T lymphocyte response to foreign antigen / self major histocompatability (MHC) complexes. The pathways responsible for generating DCs in vivo remain to be clearly defined. However, in vitro, dendritic cells can be

20 derived from the myeloid (Inaba et ah, 1992; Scheicher et ah, 1992) or lymphoid (Steinman et ah,

1997; Wu et ah, 1996) lineages, resulting in either the interstitial or Langerhans DCs in the case of the myeloid progenitor, or plasmacytoid DCs from the lymphoid lineage (Fig. 1.2). Recent evidence suggests that the myeloid and lymphoid DC subsets play differing roles in the immune response.

The lymphoid (plasmacytoid) DCs have been proposed to play a regulatory role, containing immune responses (Austyn, 1998; Shortman and Caux, 1997). In contrast myeloid DCs have a stimulatory role on an immune response. The myeloid DCs are located at anatomical sites where many pathogens enter the body and are initially exposed to foreign agents at the site of infection.

They undergo critical activation and maturational changes, resulting in the release of inflammatory cytokines and chemokines that recruit inflammatory cells to these sites. These mature myeloid DCs then transport the antigen to the T cell zone of the draining lymph nodes where they present antigenic peptides in the context of MHC class II molecules to antigen-specific T cells, thereby initiating an adaptive immune response. While there have been many recent advances in the understanding of the DC, there are still many unanswered questions about the signals required for this important maturation process.

21 CD34+ Progenitor CD34+ myeloid lymphoid

CD14+ CD14 CD14- Blood CD11C+ CD110+ CD11C- precursor CD1- CD1- IL3Roc+

Monocytes

GM-CSF GM-CSF IL-4 IL-3 Differentiating IL-4 cytokines: ^ r 1 , t g f -p I Interstitial DC Langerhans DC Plasmacytoid DC

Immature cell

Maturation CD40L signals: LPS CD40 L CD40L i MCM

Mature cell

Adapted from; Immunobiology of Dendritic cells Banchereau et al., (2000) Ann. Rev. Immune. 18, 767-811

Figure 1.2 Three subsets of human dendritic cells Myeloid CD34+ progenitors differentiate into monocytes (CD 14+ CD 11 c+ DC precursors) that give rise to immature interstitial DCs in response to GM-CSF and IL-4. Myeloid progenitors also differentiate into CD 14- CDl lc+ precursors, which give rise to Langerhans cells in response to GM-CSF, IL-4 and TGF-p. Lymphoid CD34+ progenitors differentiate into immature DCs in response to IL-3. The immature dendritic cells differentiate into mature cells in response to maturation signals such as CD40L, LPS and MCM.

22 1.2.3 Osteoclasts

Osteoclasts are formed from myelomonocytic haematopoietic lineage cells that are recruited from the blood to the bone surface, where they undergo further differentiation and fusion to form multinucleated cells. Polarisation of the osteoclast occurs on contact with bone and in response to osteoclast-inducing cytokines (to be discussed below); the cells undergo rearrangements of the actin cytoskeleton and formation of a tight sealing zone between the bone surface and the basal membrane to form a specialized compartment. Within this resorption pit (Howship’s lacunae) osteoclasts secrete acids and lytic enzymes that solubilise the bone mineral and collagen matrices and through this process the osteoclast erodes the underlying bone (Fig. 1.3). The bone resorption activity of osteoclasts is essential for bone remodelling which occurs throughout adult life. A balance between osteoblastic bone formation and osteoclastic bone resorption is critical, since imbalances of remodelling result in skeletal diseases such as osteoporosis and rheumatoid arthritis.

These diseases are mostly caused by excess osteoclastic activity and, as such, the osteoclast and the mechanisms of bone resorption have become major targets for therapeutic intervention.

The differentiation of osteoclasts in vitro, utilizing mouse bone marrow cultures of spleen cells and

stromal cells has led to the discovery of RANKL and macrophage-colony stimulating factor (M-

CSF) as the two factors necessary for osteoclast differentiation (Yasuda et al., 1998). RANKL

induces the activation of NF-kB and, together with M-CSF, induces the expression of genes

associated with the osteoclast lineage, such as TRAP, calcitonin receptor, cathepsin K and the |33

integrin (Lacey et al., 1998). The importance of RANKL for osteoclastogenesis came initially from

the discovery of osteoprotegrin (OPG), a soluble receptor that prevents RANKL from binding the

RANK receptor on osteoclast precursor cells. Mice deficient in RANK or RANKL are

osteopetrotic, with fragile bones (Kong et al., 1999c; Li et al., 2000; Simonet et al., 1997) as are

mice over expressing OPG, the naturally occurring soluble inhibitor of the RANK / RANKL

23 interaction (Li et al., 2000). The importance of M-CSF in bone was demonstrated by the discovery of a mutation in the M-CSF gene in the op/op (osteopetrotic) mutant mouse (Marks, Jr. and Lane,

1976).

Adapted from; Modulation of osteoclast differentiation, Suda et al., (1992) Endocrine Reviews 13, 66-79

RANKL RANK

Calcitonin

HXO Calcitonin receptor Chloride Channel ysosome Proton Ruffled border TRAP Pump

ly^so m a l enzyme

Figure 1.3 A schematic representation of osteoclast molecules and functions The osteoclast attaches to bone and, in the presence of RANKL, formation of the ruffled border occurs. Numerous aVP3 molecules and receptors for calcitonin and RANKL are expressed on the osteoclast plasma membrane. Osteoclasts possess high levels of TRAP activity, which is conveniently used as a specific histochemical marker. A resorptive microenvironment is achieved beneath the cell by the synthesis of hydrochloric acid from the hydrogen ions provided by proton pumps and secretion of chloride ions into the resorption pit via chloride channels. Lysosomal enzymes are secreted into the resorbing area and facilitate degradation of the organic bone matrix.

24 1.2.4 Macrophages, dendritic cells and osteoclasts

Perhaps osteoclasts and myeloid dendritic cells represent specialised cell types displaying developed but mutually exclusive features of macrophages. Indeed, early reports of collagen and bone resorption by osteoclasts referred to this as phagocytosis (Everts et al., 1985), a fundamental property of macrophages. Furthermore, activated macrophages are able to degrade the osteoid covering the bone surface (Lassus et al., 1998) and macrophages are capable of antigen presentation

(Unanue, 1984), but unlike dendritic cells, macrophages do not have the capacity to activate naïve T cells (Liu and MacPherson, 1993).

There is increasing evidence for a common precursor for myeloid dendritic cells and osteoclasts.

Murine bone marrow Flt3-positive monocyte precursors have been demonstrated to commit sequentially to DCs and osteoclasts (Servet-Delprat et al., 2002). Each time frame corresponds to expression of a specific differentiation potential: day 6 monocyte precursors show preference for osteoclast differentiation while day 8 precursors have a preference for DC differentiation. The bipotential of common precursor cells to differentiate into osteoclasts and dendritic cells has been demonstrated at the single cell level (Miyamoto et al., 2001). In these studies, dendritic cell differentiation was inhibited using M-CSF, a co-factor for osteoclastogenesis; in contrast, osteoclast differentiation was inhibited by granulocyte macrophage-colony stimulating factor (GM-CSF), a factor essential for dendritic cell differentiation. The authors were also able to demonstrate concomitant regulation of the proto-oncogene associated with commitment to the osteoclast lineage

(Grigoriadis et al., 1994). The use of in vitro differentiation systems to derive dendritic cells and osteoclasts from a common precursor cell should enable investigation of phenotypic changes that occur during the differentiation process, as well as exploration of the function of any differentially expressed molecules, perhaps yielding clues to the biology behind the very different functional capabilities of these two cell types.

25 1.3 DC isolation and differentiation pathways

Human DC can be isolated in two ways. Firstly, populations can be isolated by negative selection from peripheral blood, or enzymatically digested lymphoid tissue using a range of sophisticated flow cytometry and immunomagnetic bead systems. These systems capitalise on the absence of certain lineage-specific markers on DCs that are expressed on other defined leukocyte populations, e.g. T lymphocyte (CD3), B lymphocyte (CD 19, CD20), NK cells (CD 16, CD56), monocytes (high density CD 14) and granulocytes (CD 15) markers. The negative selection is then followed by positive selection using anti-CD83 (Zhou and Tedder, 1995), or mAbs with specificity for other

DC-specific cell-surface molecules such as BDCA-2, -3, -4 and CDlc (Dzionek et al., 2000).

Secondly, density gradient centrifugation over various gradient media such as Percoll (Amrad,

Pharmacia Biotech) (Young and Steinman, 1988) or Nycodenz (Nycomed Pharma) (McLellan et al., 1995a) exploits the low density of DCs following a period of in vitro culture. However, these media may induce phenotypic or functional changes in the cells (McLellan et al., 1995b). While ex vivo isolation of DCs is desirable for the study of DC biology, the aforementioned methods of isolation are complicated and time consuming. In addition, they may require large amounts of human blood or tissue in order to obtain sufficient numbers of cells for study. Therefore development of simple in vitro differentiation protocols for generating large numbers of functionally and phenotypically relevant DCs provides a means of gaining important insights into the mechanisms controlling the physiological generation of DCs and their functions.

Human DCs can be derived in vitro using several cellular sources: CD34-positive progenitor cells isolated from cord blood or peripheral blood, adherent peripheral blood mononuclear cells (PBMC) or peripheral blood monocytes (PBM). Cord blood cells cultured in GM-CSF and TNF-a produce a population described as dendritic Langerhans cells, with marked allo-stimulatory properties (Caux et al., 1992). Caux et al. have also demonstrated that lL-3 can substitute for GM-CSF in this system

26 (Caux et al., 1996). Furthermore, the addition of transforming growth factor-(TGF-)(3 promotes the formation of these cells (Strobl et al., 1996). Peripheral blood CD34-positive progenitors cultured in

GM-CSF (Bernhard et al., 1995) or GM-CSF plus IL-4 with a cocktail of stem cell factor (SCF), erythropoietin (EPO), IE-1(3, IL-3 and IL - 6 (Mackensen et al., 1995) gives rise to populations of immunostimulatory Langerhans cells, with varying success. Peters et al. (Peters et al., 1987) demonstrated the differentiation of a population of dendritic cells capable of antigen presentation from cultured peripheral blood monocytes with GM-CSF and IL-4. Sallusto and Lanzavecchia further demonstrated phenotypical and functional changes in this population, akin to dendritic cell maturation, on treatment with TNF-a (Sallusto and Lanzavecchia, 1994). Bender et al. showed that dendritic cells derived from lymphocyte depleted peripheral blood in the presence of GM-CSF and

IL-4 required the addition of a monocyte-conditioned medium (MCM) to give phenotypically and functionally mature DCs (Bender et al., 1996). A plethora of factors have since been shown to activate or mature in v//ro-derived immature DC populations, including LPS (Granucci et al., 1999) and CD40 Ligand (Wurtzen et al., 2001). Lastly, despite initial shortcomings in the ability of cytokine cocktails to replace monocyte-conditioned medium as a myeloid DC maturation factor

(Reddy et al., 1997), a cytokine cocktail containing GM-CSF, IL-4, IL-ip, TNF-a, IL - 6 and prostaglandin E2 (PGE2) has recently been shown to be as efficient, and potentially more reproducible, than MCM in activating immature DCs (Choi et al., 2000; Spisek et al., 2001).

The human immature DCs derived in vitro from peripheral blood monocytes in the presence of

GM-CSF and IL-4 represent a population that mimic the common characteristics of a DC (Fig. 1.4).

These cells express the chemokine receptors CCRl and CCR5 (Pashenkov et al., 2002; Sallusto et al., 1999). This is consistent with their requirement to migrate to areas of inflammation where ligands for these receptors, such as CCL3 (macrophage inhibitory protein- (MlP-)la), CCL4 (MIP-

IP), CCL5 (RANTES) and CCL15 (MIP-5) are expressed (Pashenkov et al., 2002; Sozzani et al.,

1997). Exposure to a variety of inflammatory or bacterial stimuli, including monocyte-conditioned

27 medium (Bender et al., 1996; Reddy et al., 1997), leads to maturation of monocyte-derived DCs.

Coordinated changes in gene expression occur upon DC maturation, including decreased expression of CCRl and CCR5 and increased expression of CCR7 (Sallusto and Lanzavecchia, 2000; Sozzani et al., 1998; Caux et al., 2000). This change in chemokine receptor expression enables the chemotactic recruitment of DCs to areas of lymph nodes abundant in T cells that express the CCR7 ligands CCL19 (ELC, MIP-3P) and CCL21 (SEC, 6 Ckine) (Baekkevold et al., 2001). DC maturation also results in increased expression of many proteins other than chemokines, such as the co-stimulatory molecules CD40, CD80 (B7.1) and CD 8 6 (B7.2) (Fig. 1.4), fundamental to the ability of DCs to interact with and activate naïve T cells (Ni and ONeill, 1997). The fact that

CD40L expressed by activated T cells is used as a potent maturation factor for DCs, demonstrates the importance of two-way communication between DCs and T cells. The maturation process further affects some of the functional properties of DCs (Fig. 1.4). Immature DCs demonstrate efficient antigen uptake, whereas mature DCs show a reduced ability to endocytose antigen and an increased ability to present peptide / MHC complexes to naïve T cells. Although the terms immature and mature are used for these cell types, with the mature cells capable of being induced from the immature population (Reddy et al., 1997), relevant terminology such as activated or primed could also be used to describe mature DC.

28 IL-4 GM-CSF MCM

Immature DC M ature DC

Antigen uptake +++ + Antigen Processing +++ + MHC class II + +++ Costimulatory molecules + +++ T cell activation +++ Chemokine Receptors C C R 1, C C R 5 C C R 7

Figure 1.4 The differentiation and maturation of m vitro monocyte-derived DCs DCs acquire specific abilities upon maturation and lose certain others relevant to their roles at each stage of the differentiation process. Immature DCs have powerful antigen uptake and processing capabilities, express MHC class II and low levels of co-stimulatory molecules consistent with their ability to activate antigen-specific T cells. Mature DCs have lost the antigen uptake and processing capacity and have increased MHC class II and co-stimulatory molecule expression, allowing more efficient activation of antigen-specific T cells.

29 1.4 Osteoclast isolation and differentiation pathways

Several techniques have been developed for isolating osteoclasts from long bones. Commonly, osteoclasts are prepared from embryonic chicken tibia by removing the marrow and releasing the osteoclasts using calcium-free and magnesium-free buffers. Enrichment of multinucleated osteoclasts then occurs, by sieving through a nylon mesh, followed by density gradient centrifugation (Osdoby et al., 1982). Isolation of osteoclasts from juvenile rabbit long bones recovers large numbers of osteoclasts (Tezuka et al., 1994). However, these complicated methods result in heterogeneous populations with variable bone resorption capabilities.

Mouse bone marrow culture systems have been established in which Tartrate Resistant Acid

Phosphatase (TRAP)-positive multinucleated cells form in 5 to 6 days and are capable of bone resorption (Takahashi et al., 1988). These methods have been extended to human bone marrow cultures, with the addition of 1,25-dihydroxyvitamin D3, in which osteoclast cells develop in 3 weeks (MacDonald et al., 1987; Thavarajah et al., 1991). Co-culture systems then started to emerge where mouse spleen cells were used as a source of osteoclast precursors that formed multinucleated cells on culture with bone marrow-derived stromal cell lines (Udagawa et al., 1989). These co­ cultures can be established on collagen-coated plates from which osteoclast-like cells can be released by collagenase (Akatsu et al., 1992). The immortalisation of osteoclast precursors has resulted in cell lines capable of generating high numbers of bone resorbing osteoclasts (Hentunen et al., 1998; Chen and Li, 1998).

Since the description of RANKL, a TNF superfamily member capable of replacing stromal cells in co-culture systems to induce osteoclast formation (Yasuda et al., 1998), in vitro differentiation systems using precursor cells on their own have been developed. Treatment of adherent human

PBMCs with RANKL and M-CSF induces TRAP-positive, multinucleated osteoclast-like cells expressing aVp3 and calcitonin receptors (CTRs) (Matsuzaki et al., 1998; Quinn et al., 1998). The

30 formation of osteoclast-like cells from the culture of CD 14-positive mononuclear cells with

RANKL and M-CSF, confirms these peripheral blood monocytes as the osteoclast precursor contained in the PBMC fraction (Massey and Flanagan, 1999). Indeed, enhanced osteoclast formation is observed when the adherent fraction of PBMC is used for osteoclast cultures.

Otherwise, TGF-(3 is required to restore osteoclast differentiation in the presence of the non­ adherent fraction to a level equivalent to using the adherent fraction alone as a precursor population

(Massey et al., 2001).

Culturing of human peripheral blood monocytes in the presence of M-CSF and RANKL gives rise to a population of cells that have all the characteristics of osteoclasts. They are multinucleated,

TRAP-positive, CTR-positive, aVpS-positive and, most importantly, are capable of bone resorption

(Fig. 1.5). This, therefore, represents a system amenable to the further study of osteoclast generation and function.

31 RANKL M-CSF RANKL m

m onocyte pre-osteoclast mature o steo cla st

TRAP-negative TRAP-positive TRAP-positive CTR-negative CTR-positive CTR-positive aVp3-negative aVp3-positive aVp3-positive single nuclei single nuclei multinucleated no resorption no resorption bone resorption

Figure 1.5 Schematic for the in vitro differentiation of osteoclasts and characteristics of these cells M-CSF and RANKL are essential for the differentiation of osteoclasts from peripheral blood monocytes. During this differentiation process TRAP is expressed and CTR and aV(33 are acquired. Pre-osteoclasts fuse in the presence of RANKL to form multinucleated osteoclasts. These multinucleated osteoclasts are capable of bone resorption.

32 1.5 Gene expression: mRNA transcript analysis

The aim of this project was to analyse the expression of a selected set of genes in differentiated osteoclasts, differentiated dendritic cells and their monocyte precursors. Each cell type has a unique pattern of gene expression, the analysis of which can reveal the underlying physiologies of the cell.

There are several ways of performing such analysis. The most straightforward way is to construct a representative cDNA library from the cell type of interest and sequence every cDNA clone. This is obviously a monumental task. While this approach is probably the most objective, a more focussed approach around gene families may reap greater rewards in the short term. To analyse the expression of specific genes, mRNA can be isolated from the cells and subjected to reverse transcription-PCR (RT-PCR) using a panel of specifically designed oligonucleotide primer pairs

(two short single-stranded DNA fragments with sequences complementary to those that flank the stretch of cDNA to be amplified). Detection of specific RNA sequences can be confirmed by

Northern blotting in which RNA isolated from cells is electrophoretically separated by size in the presence of a denaturing agent, such as formaldehyde, and then transferred to a solid-support nitrocellulose membrane. Hybridisation of specifically designed probes (labelled radioactively for example) to the membranes allows specific RNA sequences to be identified. However this process is best suited to analysis of a small number of defined genes. Where a large numbers of genes are to be investigated, comparative gene expression profiling is used.

One application of RT-PCR is a method for differential mRNA screening called differential display

RT-PCR (Liang and Pardee, 1998). In this technique, two or more cell populations can be compared and mRNA is isolated from each population. Reverse transcription of mRNA into single strand cDNA followed by PCR is performed in the presence of a source of radiolabelled nucleotide using an oligo-dT primer, which will anneal to the 3' poly A tail of all mRNA species, and a set of random primers (usually lOmers), which will anneal to complimentary sequences upstream of the

33 poly A tail. Since the upstream primer will anneal to different cDNA species at different relative positions, the lengths of the PCR products will vary for multiple mRNA species detected. The products from these reactions are then separated on a high-resolution gel and exposed to X-ray film.

The presence of bands with a greater intensity in one lane compared with another represents increased expression of an mRNA species in one cell population compared with another. The cDNA can be subsequently recovered from the gel for further analysis and identification.

Another technique termed serial analysis of gene expression (SAGE) involves sequencing a small and unique fragment of each expressed gene (called a SAGE tag). The number of times this fragment appears (the SAGE tag number) directly reflects the abundance of the corresponding transcript. The generation of a SAGE cDNA library does not require any prior knowledge of what genes are expressed in the cell of interest and is also able to detect and quantify the expression of previously uncharacterized genes. SAGE is based on two fundamental principles; since a 10 base pair sequence has 4*° possible combinations and there are aproximately 30,000 genes within the human genome, a short (10-11 bp) oligonucleotide fragment (SAGE tag) is sufficient to uniquely identify any transcribed human gene. Secondly, multiple 10 base pair SAGE tags can be concentrated in one plasmid, meaning that a single sequencing reaction can provide information for

30 to 35 different genes.

A further approach for comparing gene expression profiles is the use of DNA microarrays. Two basic types of DNA microarrays are commonly used: oligonucleotide arrays and cDNA arrays. Both approaches involve the immobilization of DNA sequences on a girded array using a solid support surface, such as a glass microscope slide or nitrocellulose membrane. In the case of oligonucleotide arrays, fragments of known DNA sequence, typically 20 to 25-nucleotides in length are synthesized in situ on the surface of the support, allowing as many as 300,000 distinct sequences representing over 6,000 genes on a single 1.3cm x 1.3cm microarray. In the case of cDNA microarrays, PCR amplified cDNA fragments are deposited onto the solid support using a robotic spotting device. For

34 both microarray approaches, the next step involves the purification of RNA from the source of interest (for example, in vfrro-derived cells), labelling of the RNA, and hybridisation of the labelled material to the microarray. Hybridisation events are then captured by scanning the surface of the microarray, with the intensity of each spot on the array proportional to the level of expression of the gene represented by that spot (outlined in Fig. 4.1). This technology provides the ability to monitor the expression levels of thousands of genes simultaneously.

1.6 mRNA transcript analysis of a selection of genes in myeloid cells

DNA microarray technology could be used to evaluate the differential expression of candidate genes in lineage related dendritic cells and osteoclasts and their monocyte precursors. G protein coupled receptors have emerged as excellent targets for therapeutic intervention, that is, greater than

30% of all currently marketed pharmaceuticals are modulators of specific GPCRs (Wise et al.,

2002). The defined distribution and function of a subset of these receptors (chemokine receptors) and their ligands (chemokines) in one of the myeloid cell populations, namely dendritic cells, led to the proposal that differential expression of G protein coupled receptors and their ligands in osteoclasts and dendritic cells could bring new knowledge to the biology of these cell types and help to identify important new therapeutic targets. Expression of chemokine and chemokine receptor genes, related G protein-coupled seven transmembrane receptors and their ligands were therefore evaluated in immature and mature DCs, osteoclasts and their monocyte precursors using a custom-made DNA microarray. The following passages introduce the gene families represented on the microarray and their significance in dendritic cell and osteoclast biology.

1.7 The GPCR superfamily

Seven-transmembrane (7TM) receptors, or G protein-coupled receptors (GPCRs), form the largest gene superfamily of cell-surface receptors. GPCRs, also known as Serpentine Receptors are plasma

35 membrane proteins that span the membrane seven times, containing an extracellular NH 2 -terminus and a cytoplasmic COOH-terminus. More than 2000 GPCRs have been reported since the discoveries of bovine opsin (Nathans and Hogness, 1983) and the (3-adrenergic receptor (Dixon et ah, 1986). GPCRs are grouped into four defined classes; receptors related to rhodopsin (class A), those related to secretin receptors (class B), receptors related to the metabotropic receptors (class C) and the frizzled receptors. Other less defined families include the fungal pheromone receptor family

(class D) and the cyclic adenosine monophosphate (cAMP) receptors (class E) (G protein-coupled receptor database, www.gpcr.org). Each class of GPCRs shows substantial differences in their transmembrane sequences and cannot be traced to a single evolutionary origin.

The central role of GPCRs in multicellular organisms is reflected by their diversity. GPCRs respond to a wide range of stimulants including light, hormones, nucleotides and odorants (Pierce et al., 2002). Ligand binding occurs at varied sites on these receptors, including sites within the hydrophobic core formed by the transmembrane a-helices, within the NH 2 -terminus and extracellular hydrophobic loops that link the transmembrane domains (Gether and Kobilka, 1998).

Ligand occupancy induces a conformational change in the GPCR that recruits and activates various

G-proteins which in turn stimulate the generation of cAMP, phosphoinositides, diacylglycerol and other second messengers. This then triggers such events as activation of kinase cascades and phosphorylation of cytosolic factors and nuclear transcriptional factors (Brivanlou and Darnell, Jr.,

2002).

Modulation of GPCR function has resulted in a wide range of important therapeutic agents. Thus, the identification of new GPCR ligands, may lead to important discoveries in human physiology.

There are numerous orphan GPCRs, representing new families, possibly with distinct and novel ligands (Wilson et al., 1998). Recent advances in 7TM receptor biology include many successes in the pairing of orphan 7TM receptors with their cognate ligands. However there are few remaining endogenous ligands whose receptor partners have not yet been identified and many receptors

36 remain orphans, prompting efforts to find previously unknown ligands that might activate 7TM receptors (Howard et ah, 2001). Recent research indicates receptor-protein interactions demonstrating 7TM receptor signalling via G protein independent mechanisms, such as the angiotensin lA receptor interacting through its carboxyl tail with the Janus kinase, JAK2, leading to transcription factor activation (Marrero et al., 1995). Furthermore the first 7TM receptor, rhodopsin, has now been crystallised (Palczewski et al., 2000), allowing for the evaluation of receptor / ligand interactions at a molecular level.

1.8 Chemokine receptors

There is a vast literature encompassing chemokines and chemokine receptors. Table 1.1 attempts to summarise the current knowledge about chemokines, chemokine receptors and their presumed functions. Chemokine receptors belong to the class A (rhodopsin like) GPCR subfamily. These chemokine receptors are classified into four families: XCR, CCR, CXCR and CX3CR, binding to the C, CC, CXC and CX3C chemokines respectively. To date, four CXC chemokine receptors

(CXCRl - CXCR4), ten CC chemokine receptors (CCRl - CCRIO) and one CX3C chemokine receptor (CX3CR) have been cloned and characterized. Most known chemokine receptors are reported to bind to several ligands of the same class and most chemokines can interact with more than one receptor, generating considerable redundancy within these families of molecules. In addition to chemokine receptors with known ligand(s), numerous orphan receptors that look to be members of the chemokine receptor family have been reported.

1.8.1 CXC receptors

CXCRl and CXCR2, previously known as IL- 8 RA, or type IIL - 8 receptor, and IL- 8 RB, or type II

IL- 8 receptor, share 77% amino acid sequence identity. CXCL 8 (IL-8 ) binds to both receptors with high affinity and induces rapid elevation of cytosolic Ca^^ levels (Murphy and Tiffany, 1991; Lee et

37 al., 1992; Holmes et al., 1991). CXCR2 has been shown to bind with high-affinity to other Glu-Leu-

Arg (ELR) motif containing chemokines including CXCLl, CXCL2, CXCL5, CXCL 6 and CXCL7.

In contrast, CXCL4, CXCL9 and CXCLIO, CXC-chemokines that lack the ELR motif, have been shown not to bind CXCRl and CXCR2 (Clark-Lewis et al., 1993). CXCRl and CXCR2 are expressed by neutrophils, but not B lymphocytes or T lymphocytes. CXCR3, also known as the IP-

10 / Mig receptor, shares 40% protein sequence identity with CXCRl and CXCR2 (Loetscher et al.,

1996). CXCR3 is highly expressed by IL-2-activated T lymphocytes but is absent from resting T lymphocytes, B lymphocytes, monocytes and granulocytes. CXCR3 binds CXCL9 and CXCLIO with high affinity and mediates Ca^^ mobilization and chemotaxis. CXCR3 does not bind CXCL4 or any of the CXC chemokines containing the ELR motif. CXCR4 previously termed fiisin or

LESTR was originally discovered as an orphan receptor with structural similarity to chemokine receptors (Feng et al., 1996; Loetscher et al., 1996). CXCR4 was subsequently identified as a necessary cofactor for entry of T cell-tropic HIV viruses into CD4-positive T cells (Oberlin et al.,

1996). The CXC chemokine CXCLl2 (SDF-1) has now been shown to be the ligand for CXCR4 and a powerful inhibitor of infection by T cell-tropic HIV-1 strains (Bleul et al., 1996; Oberlin et al., 1996).

1.8.2 CC receptors

CCRl was the first identified CC chemokine receptor and is expressed on monocytes, neutrophils and eosinophils. CCRl binds CCL3 (MIP-la), CCL5 (RANTES) and CCL7 (MCP-3) with high affinities and also CCL2 (MCP-1) and CCL4 (MIP-ip) with lower affinities (Neote et al., 1993;

Gao et al., 1993). CCR2 binds all four members of the Monocyte Chemoattractant Protein (MCP) family, CCL2, CCL8 , CCL7 and CCL13 also known as MCP-1-4 respectively; CCL7, CCL 8 and

CCL13 also interact with CCRl and CCR5 (Baggiolini et al., 1997; Ruffing et al., 1998). CCR2 is expressed on monocytes but not on human neutrophils or eosinophils, while CCR3 is expressed

38 predominantly on eosinophils. CCR4 was originally cloned from a human immature basophilic cell line and has since been shown to be expressed in T cells and IL-5-primed basophils. CCR4 has been shown to mediate the biological activities of CCL2, CCL3 and CCL5. CCR5 is expressed on primary adherent monocytes, but not on neutrophils or eosinophils. CCR5 mediates the activities of

CCL3, CCL4 and CCL5. Recently CCR5 has also been shown to be a co-receptor on CD4-positive target cells for infection with primary monocyte-tropic human immunodeficiency virus-1 (HIV-1) viruses. CCR 6 is expressed on activated T cells and dendritic cells and binds CCL20 (MIP-3a)

(Baba et al., 1997; Greaves et al., 1997). CCR7 has been shown to bind the chemokines CCL19

(MIP-3p) and CCL21 (SLC) and is involved in homing of T cells and DCs to lymphoid tissues during a primary immune response (Yoshida et al., 1998a; Yoshida et al., 1998b). CCR 8 mediates the biological activities of CCLl (1-309), CCL4 and CCL17, demonstrates lymphoid tissue restricted expression and there is evidence that it is up-regulated in Th2-type T lymphocytes, suggesting a role in lymphocyte migration, activation and differentiation (Napolitano and Santoni,

1999). CCR9 currently has only one known ligand, CCL25 (TECK), whereas CCRIO has several ligands including CCL27 (CTACK) and CCL28 (MEC). CCL25, CCL27 and CCL28 are all epithelial-expressed and prevalent in the small intestine, skin and mucosa respectively. As such a role for CCR9 and CCRIO in lymphocyte homing to these peripheral sites has been postulated

(Kunkel and Butcher, 2002).

1.9 Chemokines

Chemokines are a family of small structurally related molecules with a major role in regulating leukocyte recruitment and mediate their activities through binding to chemokine receptors. There are four classes of chemokines, defined by the arrangement of the conserved cysteine (C) residues of the mature proteins: the CXC chemokines have one amino acid residue separating the first two conserved cysteine residues; the CC chemokines in which the first two conserved cysteine residues

39 are adjacent; the C chemokine class lacks two (the first and third) of the four conserved cysteine residues and the CX3C chemokine category, in which the cysteines are separated by three residues.

Within the CXC subfamily, the chemokines can be further divided into two groups. One group of the CXC chemokines have the characteristic three amino acid sequence ELR (glutamic acid- leucine-arginine) immediately preceding the first cysteine residue near the NH 2 -terminus. A second group of CXC chemokines lacks this ELR motif. The CXC chemokines containing the ELR domain, including CXCLl (GRO-a), CXCL2 (GRO-P), CXCL3 (GRO-y), CXCL5 (ENA-78)

CXCL6 (GCP2), CXCL7 (NAP-2) and CXCL 8 (IL-8 ), act primarily on neutrophils as chemoattractants and activators, inducing neutrophil degranulation with release of myeloperoxidase and other enzymes. The CXC chemokines without the ELR domain, such as CXCL4 (PF4), CXCL9

(MIG) and CXCLIO (IP 10) and the C chemokines XCLl (lymphotactin) and XCL2 are chemoattractive for lymphocytes. The CC chemokines, such as CCL2 (MCP-1), CCL3 (MIP-la),

CCL4 (MIP-iP) and CCL5 (RANTES) chemoattract and activate monocytes, dendritic cells, T lymphocytes, NK cells, B lymphocytes, basophils and eosinophils to varying degrees. The only

CX3 C chemokine CX 3CLI (ffactalkine), unlike other known chemokines, is a type I membrane protein containing a chemokine domain at the NH 2 -terminus tethered on a long mucin-like stalk, which has been shown to be involved in adhesion either as the membrane-tethered form or as a proteolytically shed form (Rossi and Zlotnik, 2000) and is expressed predominantly by epithelial cells (Lucas et al., 2001).

In addition to their roles in regulating leukocyte recruitment and trafficking, certain chemokines have been reported to act on haematopoietic progenitor cells and on non-leukocyte cells such as fibroblasts, smooth muscle cells, kératinocytes and melanoma cell lines. Other chemokines have also been implicated as playing a role in wound healing, in angiogenesis and in viral infection. In in vitro assays, chemokines have overlapping and redundant functions. It remains to be determined to

40 what extent the various chemokines have unique roles in vivo. To date, over 45 chemokines have been described in humans (Nelson and Krensky, 2001). The majority of genes for all CC chemokines and CXC chemokines are grouped on two distinct chromosomal loci, 17ql 1.2-12 and

4ql2-21 respectively (Naruse et al., 1996). The clustering of chemokine genes suggests that many chemokine family members arose through gene duplication and subsequent sequence divergence.

Even the exceptions, CCL17, CCL22 and CX 3 CLI, which are chromosome 16ql3 chemokines, demonstrate synchronised regulation by way of their similar expression patterns in response to Th2- type cytokines and in atherosclerotic plaques (Greaves et al., 2001).

Chemokines can be further divided into two groups, inflammatory and homeostatic or homing chemokines. Inflammatory chemokines are upregulated in peripheral tissues in response to inflammatory cytokines, bacterial toxins and other inflammatory mediators. The ligands for CCRl,

CCR2, CCR3, CCR5, CXCRl, CXCR2 and CXCR3 belong to this category. Inflammatory mediators do not generally induce expression of the homeostatic chemokines; rather, these chemokines are produced constitutively within defined areas of lymphoid tissues. The main fimction of the homeostatic chemokines is to control the homing of lymphocytes to the lymphoid tissues. In general inflammatory chemokines display an element of redundancy, with their receptors responding to multiple ligands, whereas the receptors for homing chemokines such as CCR7, CCR9 and CXCR4 are selective for one, or at the most two, ligands and their expression is restricted to lymphocytes and mature dendritic cells.

41 SYSTEMIC OTHER NAMES RECEPTOR MAIN FUNCTIONS NAME USAGE CXCLl GRO-oc/MGCA-a CXCRl, CXCR2 Neutrophil migration CXCL2 GRO-p/MGSA-p CXCR2 Neutrophil migration CXCL3 GRO-y/MGSA-y CXCR2 Neutrophil migration CXCL4 PF4 Unknown CXCL5 ENA-78/LIX CXCR2 Neutrophil migration

CXCL6 GCP-2 CXCRl, CXCR2 Neutrophil migration CXCL7 NAP-2 CXCR2 Neutrophil migration

CXCL8 IL- 8 CXCRl, CXCR2 Neutrophil migration CXCL9 Mig CXCR3 T cell migration CXCLIO IPIO CXCR3 T cell migration CXCLl 1 I-TAC CXCR3 T cell migration CXCL12 SDF-1 o/p CXCR4 B cell lymphopoiesis CXCL13 BLC/BCA-1 CXCR5 B cell trafficking CXCL14 BRAK/bolekine Unknown CXCLl 5 Unknown CXCL16 SexCkine CXCR6 T cell migration

CCLl 1-309 CCR8 T cell trafficking CCL2 MCP-1 CCR2 T cell and monocyte migration CCL3 MIP-loc/LD78a CCRl, CCR5 T cell and monocyte migration CCL4 M lP-lp CCR5 T cell and monocyte migration CCL5 RANTES CCRl, CCR3, CCR5 T cell, granulocyte; monocyte migration

CCL6 Unknown CCL7 MCP-3 CCRl, CCR2, CCR3 T cell, granulocyte; monocyte migration

CCL8 MCP-2 CCR3 T cell and granulocyte migration CCL9/10 CCRl C C Lll Eotaxin CCR3 T cell and granulocyte migration CCL12 CCR2 T cell and monocyte migration CCLl 3 MCP-4 CCR2, CCR3 T cell, granulocyte; monocyte migration CCL14 HCC-1 CCRl T cell and monocyte migration CCLl 5 HCC-2/Lkn-l/MIP-lô CCR1,CCR3 T cell, granulocyte; monocyte migration

CCLl 6 HCC-4/LEC CCRl T cell and monocyte migration CCL17 TARC CCR4 T cell and monocyte migration

CCLl 8 DC-Kl/PARC/AMACl Unknown CCLl 9 MlP-3 p/ELC/exodus-3 CCR7 T cell and dendritic cell migration CCL20 MIP-3 ot/LARC/exodus-1 CCR6 Dendritic cell migration CCL21 6kine/SLC/exodus-2 CCR7 T cell and dendritic cell migration CCL22 MDC/STCP-1 CCR4 T cell and monocyte migration CCL23 CKP8/MPIF-1 CCRl T cell and monocyte migration CCL24 MPIF-2/Eotaxin-2 CCR3 T cell and granulocyte migration CCL25 TECK CCR9 ^ T cell homing to gut CCL26 Eotaxin-3 CCR3 T cell and granulocyte migration CCL27 CTACK/ILC CCRIO T cell homing to skin CCL28 MEC CCR3, CCRIO T cell homing and granulocyte migration

XCLl Lynphotactin/SCM-1 a XCRl T cell trafficking XCL2 SCM -ip XCRl T cell trafficking

CX3 CLI Fractalkine/Neurotactin CX3 CRI T cell and NK cell trafficking

Table 1.1 Human chemokines, their receptors and main functions

42 1.10 Chemokines and their receptors in DC biology

The ability of DCs to both sample the microenvironment within peripheral tissues, and migrate to the lymph nodes, where they present antigen to T cells, requires a tight degree of control. The coordinated chemokine receptor expression is central to the dual role of these cells. Dendritic cells in their immature state are recruited from the blood into the tissues, via the chemokine receptors

CCRl and CCR5, in response to gradients of their ligands such as the inflammatory chemokines

CCL2, CCL3, CCL4 and CCL5 (Banchereau et al., 2000; Pashenkov et al., 2002). CCR 6 expression by immature DCs (Langerhans cells only) allows responsiveness to the inflammatory chemokine

CCL20 (MIP-3a), consistent with the prevalence of this chemokine in inflamed skin and the role of

Langerhans cells in surveillance of the skin (Greaves et al., 1997; Power et al., 1997). Exposure to inflammatory signals, such as lipopolysaccharide (LPS) and TNF-a, results in a loss of expression of these receptors for inflammatory chemokines and a corresponding upregulation of CCR7

(Sallusto et al., 1998). These mature dendritic cells therefore become responsive to the CCR7 ligands CCLl9 (MIP-3p) and CCL21 (SLC), and as such are attracted to lymphoid organs where these T cell attractive chemokines are expressed.

In addition to the tightly regulated expression of chemokine receptors in immature and mature DCs, these cells also demonstrate distinct profiles of chemokine production. In response to inflammatory stimuli immature DCs transiently produce large amounts of the inflammatory chemokines CCL3,

CCL4 and CXCL 8 , with more sustained production of CCL2, CCL5 and CXCLIO consistent with the requirement of immature DCs to migrate to areas of inflammation (Pashenkov et al., 2002), where these inflammatory chemokines will play a central role in calling in the cells involved in the ongoing immune response. Following DC maturation the production of CCLl7 (TARC), CCLl9 and CCL22 (MDC) are upregulated (Sallusto et al., 1998). These homeostatic chemokines aid the recruitment of further DCs and T cells to the T cell rich areas of the lymph nodes (Baekkevold et

43 al., 2001). Following migration to this environment, DCs and naïve T cells can interact, thereby initiating an immune response.

1.11 Chemokines and their receptors in osteoclast biology

Little is known of the importance of chemokines and their receptors in osteoclast function. What is known is that chemokines mediate activities on osteoclast precursors. For instance CCL3 and CCL4 have been shown to enhance osteoclast formation from precursor cells in vitro and the high levels of

CCL3 and CCL4 expressed in the bone marrow of patients with multiple myeloma correlate well with increased bone destruction in these patients (Abe et al., 2002). Yu et al. demonstrated CXCR4 expression from murine osteoclast precursors and subsequently showed CXCLl2 (SDF-1 )-mediated chemoattraction and transendothelial migration of these precursors (Yu et al., 2003). Other CC chemokines, such as CCL5 (RANTES) and CCL23 (CKP- 8 ), mediate chemotactic responses in osteoclast precursors; indeed CCL23 has been detected at the site of bone turnover, with CCL23 mRNA transcripts found in osteoblasts by in situ hybridization and CCL23 protein detected in osteolytic tissue by immunolocalisation (Votta et al., 2000). CXCL 8 can stimulate osteoclast formation even in the presence of excess soluble RANK receptor (RANK-Fc), suggesting a mechanism independent of RANKL induced osteoclastogenesis. Moreoover, these CXCL 8 -driven effects correlated with expression of its receptor, CXCRl, on osteoclast precursors (Bendre et al.,

2003). Actively resorbing human osteoclasts also produce CXCL 8 and indeed CXCL 8 mRNA has been demonstrated in resorbing osteoclasts present in the inflamed joints of rheumatoid arthritis patients. Furthermore CXCL 8 levels increase in osteoclast cultures in response to inflammatory mediators, suggesting an autocrine role for CXCL 8 in normal bone remodelling and action as an inflammatory mediator in pathological bone remodelling (Rothe et al., 1998).

More recently, systematic approaches to identifying the chemokines and chemokine receptors involved in murine osteoclast biology have been carried out. Lean et al systematically tested for the

44 induction of chemokine and chemokine receptor mRNA in murine bone marrow co-cultures using quantitative RT-PCR with real time analysis. They correlated increases in chemokine and chemokine receptor message with formation of TRAP in their co-cultures in the presence and absence of the osteoclast differentiation factor RANKL. Their results demonstrate the expression of

CCL9 (MlP-ly), CCL22 (MDC), CXCL13 (BLC) and CCL25 (TECK) by RANKL, with CCRl,

CCR3 and CXCRl the chemokine receptors expressed. CCL9 and CCRl were identified as the major chemokine ligand and receptor expressed by osteoclasts (Lean et al., 2002). CCL9 represents a CC chemokine with no known human homologue, however many other CC chemokines mediate their actions through CCRl and as such one of these may substitute for CCL9 in human osteoclast biology. A systematic approach to identify chemokine and chemokine receptor profiles in human osteoclasts or osteoclast like cells has not yet been reported.

1.12 Other GPCRs and ligands: an overview

The other GPCRs included in the custom designed microarray to be utilized in this project were limited according to Celltech’s interest in particular GPCR families and indeed individual target candidates. As such this overview will be limited to these families and genes (see Appendix 1).

The protease-activated receptors (PAR 1-4) are involved in activation of platelets and are widely distributed throughout the body; as such these molecules have been implicated in inflammatory, cardiovascular, digestive and respiratory functions (Kawabata and Kuroda, 2000). Indeed, PAR-1, a receptor for thrombin, is a candidate for a cardiovascular therapeutic agent and the antithrombotic effect of PAR-1 has already been demonstrated in animal models (Chackalamannil, 2003). Trypsins and tryptases are ligands for PAR-2 and thrombin also recognises PAR-3 and PAR-4 (Kawabata and Kuroda, 2000). Elastases have also been demonstrated to regulate PAR-2 signaling (Domotor et al., 2 0 0 2 ).

45 The EDG (endothelial differentiation gene) family of proteins are receptors for the lysophospholipids, lysophosphatidic acid (LPA) and sphingosine 1-phosphate (SIP); EDG 2, EDG

4 and EDG 7 mediate activity via LPA and EDG 1, 3, 5, 6 and 8 via SIP (Chun et al., 2002). The diverse physiological activities associated with these receptors include platelet aggregation, wound healing, angiogenesis and immune modulation. Therapeutic intervention by agonising or antagonising these receptors remains a possibility in immune response mechanisms, coronary artery disease and cancer (Goetzl et al., 2002)

The receptors C3aR and C5aR mediate their activity via binding to the complement components

C3a and C5a, respectively. As such these receptors have a major role in the innate immune response. These receptors are present on cells of myeloid origin as well as other immune cells, and epithelial cells (Kohl, 2001). The orphan C3AR receptor cDNA was isolated with a probe derived from the N-formyl peptide receptor, belonging to the same family as the C3aR and C5aR and shows high homology to this family of receptors (Roglic et al., 1996).

Both EMRl and CD97 are members of the epidermal growth factor (EGF)-7TM family, comprising large extracellular regions with a variable number of EGF domains linked to a 7TM region. These molecules are closely related to another member of this family, the macrophage restricted murine

F4/80 molecule. In fact EMRl is predicted to be the human ortholog of F4/80. CD97 is expressed on a diverse array of leukocytes and has a cell surface ligand, the membrane glycoprotein CD55

(decay accelerating factor), implicating it in cellular adhesion (McKnight and Gordon, 1998)

P2 Y2 receptor belongs to the purinergic and pyrimidinergic family of receptors and responds to the nucleotide ligands UTP and ATP (Williams and Jarvis, 2000). P 2 Y2 receptor agonists, which enhance ciliary beat and increase mucociliary clearance, are being evaluated in cystic fibrosis patients (Kellerman et al., 2002).

46 A series of related orphan receptors, at the time of initiation of this study, were also included on the array. GPR105, also known as KIAAOOOl and more recently renamed as the P 2 Yi4 receptor

(Abbracchio et al., 2003), was first cloned from an immature myeloid cell line (Nomura et al.,

1994). Chambers et al. (Chambers et al., 2000) have identified Uridine 5’-diphosphoglucose (UDP- glucose) as a potent agonist ligand for GPR105, although no function has yet been attributed to the receptor. The rat and mouse orthologs for GPR105, exhibiting 81 and 83% amino acid identity respectively to the human protein, both respond to the putative ligand UDP-glucose (Freeman et al.,

2001). Nucleotides represent a family of extracellular signalling molecules with wide ranging effects. For example, ADP induces platelet aggregation, through binding the P2Yn receptor

(Hollopeter et al., 2001), and ATP mediates chemotaxis, cytotoxic responses and cytokine secretion through the P2Xy receptor (North, 2002).

1.13 Project Aims

The aim of this study is to investigate gene expression by divergent myeloid cell types, focusing on chemokines, chemokine receptors, related G protein coupled receptors (GPCRs) and their ligands, thereafter allowing further investigation of differentially expressed genes and their functional implications in dendritic cell and osteoclast biology. The hypothesis being that by using microarray technology to examine the distinct gene expression patterns of highly related, yet functionally distinct cell types, one may identify genes with functional relevance to the specific cell types.

Differentiation of osteoclasts and dendritic cells from their monocyte precursors, and the subsequent characterisation of these polarised cell populations will be presented. The change in expression of some of the chemokine and chemokine receptor genes in dendritic cells following activation / maturation will be compared to information in the scientific literature and an extensive study of production / expression of these chemokines and their receptors at the protein level will be carried out. Differential expression of chemokines and chemokine receptors in osteoclasts in

47 comparison to their monocyte precursors will also be evaluated in the hope to identity molecules with a unique role in osteoclast biology. Differential expression of other seven transmembrane receptors of interest and some of their known ligands in monocytes, osteoclasts, and immature and mature dendritic cell populations will also be evaluated.

Finally, further characterisation of two differentially expressed genes will be investigated.

Expression of GPR105 by human immature monocyte-derived DC, agonist capabilities of UDP- glucose, the putative ligand for GPR105 and the functional effects of UDP-glucose on DCs will be investigated, and possible roles for GPR105 in DC biology will be evaluated. Expression of CXCL5

(ENA-78) by osteoclast cultures will be attributed to multinucleated osteoclasts by immunocytochemistry. The functionality of this osteoclast secreted CXCL5 will be investigated and possible roles in osteoclast biology alluded to.

48 Chapter 2

Materials and Methods

49 Chapter 2

2.1 Materials

2.1.1 General reagents

Unless stated otherwise, all chemical and biological reagents were obtained from Sigma-Aldrich,

Poole, UK. All tissue culture plastic was obtained from Becton Dickinson labware, NJ, USA, with the exception of round bottomed 96-well plates which were obtained from Costar, NY, USA. All tissue culture media, additives and biological buffers were obtained from Invitrogen, Paisley, UK, unless otherwise stated. Heat inactivated foetal calf serum was obtained from Sigma-Aldrich.

2.2 Methods

2.2.1 Purification of human PBMC

PBMC were isolated from normal healthy volunteers with informed consent. Whole blood was taken by venous puncture using heparinised vacutainers (Becton Dickinson). Blood was diluted 1 in

3 using RPMI 1640 and centrifuged at 400g for 30 minutes over a Ficoll-Hypaque gradient

(Amersham Biosciences, Uppsala, Sweden). Cells at the interface were removed and washed once in relevant serum-containing medium, followed by a low speed spin to remove platelets.

2.2.2 Purification of human PBM

Peripheral blood monocytes (PBM) were purified using a VarioMACS magnetic cell sorter, LS separation columns and a monocyte isolation kit (Miltenyi Biotec, Bergish Gladbach, Germany) according to the manufacturer’s instructions. This kit allows the isolation of untouched monocytes by depletion of non-monocytes with a cocktail of hapten-conjugated CD3, CD7, CD 19, CD45RA,

50 CD56 and anti-IgE antibodies. Briefly, PBMC were isolated as in Section 2.2.1, red cells were lysed by addition 20mls of red cell lysis buffer (155mM NH4CI, lOmM KHCO3, O.lmM EDTA) and rotation on a roller for 5 minutes at room temperature. The cells were then centrifuged at 200g for 5 minutes, washed in PBS and resuspended in 600|il MACS buffer per 10^ cells. 20|xl of FCR block and 20|il of hapten Ah conjugate cocktail was added per 10^ cells. Cells were mixed and incubated for 5 minutes on ice, shaking occasionally, then washed twice in lOx volume of MACS buffer

(PBS/ 0.5% BSA7 2mM EDTA). Cells were then resuspended in 600|xl MACS buffer per 10* cells.

20|il FCR block and 20|il anti-hapten microbeads were added per 10^ cells. These were mixed and incubated for 15 minutes on ice, shaking occasionally, then washed twice in lOx volume of MACS buffer and resuspended in 500|il MACS buffer. A LS+/VS+ column and adapter was located into the VarioMacs magnetic separator, the column was washed in 3ml of MACS buffer, following which the 500jxl suspension was added to the column and the eluant was collected, a further 6 ml of

MACS buffer was added to the column and eluant collected, to ensure all monocytes were eluted.

2.2.3 Monocyte-conditioned medium

Monocyte-conditioned medium (MCM) was prepared as previously described (Reddy et al., 1997).

Briefly, using Ig-coated petri dishes (100mm), 4ml of human gamma globulin (Jackson Immuno-

Research, West Grove, PA) at lOmg/ml in PBS was incubated in the petri dishes for 2 minutes. The petri dish was washed three times in PBS. 1-2 x 10^ peripheral blood monocytes were added in 5- lOmls medium. The cells were incubated for 1 hour at 37°C, and then washed with warm medium to remove non-adherent cells. 5ml medium per IdP monocytes (starting number of cells) was then added and the dishes were incubated for 24 hours at 37°C/ 5% CO 2 . MCM was collected and stored at-20°C.

51 2.2.4 Degeneration

For generation of DCs, isolated monocytes were cultured in RPMI 1640 containing 50units/ml

penicillin and 50|ig/ml streptomycin, 2mM L-glutamine and lOmM HEPES with 5ng/ml GM-CSF

and 50ng/ml IL-4 (R&D Systems, Minneapolis, MN) at 1.5 x 10^ cells/well in 6 -well tissue culture

plates for 6 days at 37°C/ 5% CO2 . Cytokines were replenished on day 3 by adding 0.5ml of fresh

medium with 30ng/ml GM-CSF and 300ng/ml IL-4. On day 6 non-adherent cells were collected by

aspiration with a Pasteur pipette and either used directly for RNA preparations, or transferred to

new 6 -well plates with fresh cytokines as on day 3. To obtain mature and immature day 9 DCs, day

6 DCs were cultured for a further 72 hours either with (mature) or without (immature) the addition

of 1ml of MCM. Immature and mature DCs were characterised by flow cytometry, their capacity to

internalise FITC-dextran and their ability to induce proliferation of allogeneic T cells with

concomitant IFN-y production (see results).

2.2.5 Flow cytometry

The enrichment of monocytes from peripheral blood was evaluated using BD Biosciences

Simultest™ flow cytometry two colour reagents: CD45-FITC/ CD14-PE; CD3-FITC/ CD19-PE;

IgGi-FITC/ IgGi-PE. Dendritic cell surface markers were evaluated using FITC-conjugated

antibodies to CD 8 6 (Serotec) and CD80 (BD Pharmingen, San Diego, CA), and PE-conjugated

antibodies to CD83 (BD Pharmingen) and HLA-DR (BD Pharmingen) by two-colour flow

cytometry. Cell surface CCRl and CCR5 were determined using directly conjugated PE- and FITC-

antibodies respectively (R&D Systems). Cell surface CCR7 was determined using a three-step

technique employing an unconjugated mouse antibody to CCR7 (Insight Biotechnology, Middlesex,

UK), anti-mouse biotin (Insight Biotechnology) and streptavidin PE (Insight biotechnology).

Briefly, cells were washed with ice-cold wash buffer comprising PBS (Invitrogen) with 5% FCS

and 0.1% sodium azide, and then incubated with optimal concentrations of unconjugated mAbs,

52 FITC- and PE-conjugated mAbs or matched isotype controls for 30 minutes on ice. Cells were washed in ice-cold wash buffer, following which cells incubated with directly conjugated antibodies were analysed immediately. Cells incubated with unconjugated antibodies were then incubated with biotin-conjugated anti-mouse IgG for 30 minutes on ice, washed extensively and finally incubated with streptavidin-PE for 30 minutes on ice, washed in ice cold buffer and analysed using a

FACScalibur equipped with CellQuest software (BD Biosciences, San Jose, CA). Live cells were gated on their forward and side light scatter properties and where indicated, the percentage of positive cells was recorded.

2.2.6 Endocytosis assays

Endocytic ability was measured by the cellular uptake of FITC-dextran (Sigma-Aldrich) analysed by flow cytometry. The DC preparations were incubated in tissue culture medium containing

0.5mg/ml FITC-dextran for 30 minutes at 37°C or washed and incubated in 0.5mg/ml FITC-dextran for 30 minutes at 4°C (control). Cells were then washed extensively in ice-cold wash buffer and the uptake of FITC-dextran was determined by flow cytometry. Results were expressed as percentage of positive cells.

2.2.7 Allogeneic T cell proliferation and IFN-yproduction

The T cell stimulatory capacities of the immature and mature DC populations were assessed by their ability to drive proliferation of naïve allogeneic T cells in a mixed leukocyte reaction (MLR).

T cells were purified from PBMC via negative selection using a pan T cell isolation kit and the

MACS column system (Miltenyi Biotec), according to the manufacturer’s protocol. The DC preparations were washed extensively and y-irradiated for 45 minutes ('*^Cs source). T cells (1x10^

cells/ml) were stimulated with 1 x 10'^ irradiated DCs/ml for 5 days at 37°C/ 5% CO 2 , in round bottomed, 96 well microtitre plates. Culture supernatants were removed on day 5 and assayed for

53 IFN-y using a Duoset™ ELISA development kit from R&D Systems, as per manufacturer’s protocol. Cellular proliferation was assessed by adding IjiCi/well of methyl-[*H]-thymidine

(lOCi/mmol; Amersham Biosciences) during the last 16 hours of culture. Cells were then harvested onto glass-fibre filter mats using a Skatron plate harvester, and radioactivity was counted using liquid scintillation.

2.2.8 IL-12production by DCs

DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4 (see Section

2.2.4). Day 6 immature DCs were incubated with MCM and culture supernatant was removed at various time points, centrifuged at 400g to obtain cell-free supernatants and stored at -20°C prior to assay. Samples were assayed for IL-12 p40 and IL-12 p70 using Duoset™ ELISA development kits from R&D Systems, as per manufacturer’s protocols. IL-12 p70 production from immature DCs was investigated further. To achieve this, day 6 immature DCs were stimulated with lOOng/ml IFN- y and/or 0.2% w/v Staphylococcus aureus Cowan’s (SAC), strain also called Pansorbin^"^

(Calbiochem, San Diego, CA) and supernatants were removed at 48 hour.

2.2.9 Osteoclast generation

PBM were diluted to 2 x 10^ cells/ml in complete medium (MEM with 15% heat inactivated FCS,

50units/ml penicillin, 50|Xg/ml streptomycin and 2mM L-glutamine) and plated out at 0.5ml/well in a 24-well plate, incubated for 45 minutes at 37°C/ 5% CO], washed once with warm MEM, aspirated and 1ml complete medium plus lOOng/ml RANKL (Peprotech EC Ltd, London, UK) and

50ng/ml M-CSF (Insight Biotechnology, Wembley, UK) was added to each well. On day 3 and 6 ,

0.5ml/well of medium was removed and replaced with 0.5ml/well medium plus 200ng/ml RANKL and lOOng/ml M-CSF. On day 9, non-adherent cells were washed away with PBS. Cells were fixed

54 and stained, used for RNA extraction or removed and re-seeded onto dentine to demonstrate formation of resorption pits.

2.2.10 Osteoclast culture staining

24-well tissue culture plates containing osteoclast cultures were washed in PBS and fixed for 20 minutes at room temperature in PBS/ 3.5% paraformaldehyde/ 2% sucrose. Plates were washed four times with PBS and either stained for the presence of TRAP, or permeabilised and stained for the presence of aVp3 and the visualisation of nuclei.

TRAP staining Paraformaldehyde fixed osteoclast cultures were stained for the presence of TRAP using the Sigma kit (387A), following the manufacturer’s instructions modified for 24-well plates. Briefly, 45ml of deionised water was warmed to 37°C. 0.5ml of fast garnet GBC base solution and 0.5ml of sodium nitrite were added to a test tube, mixed by gentle inversion for 30 seconds and allowed to stand for

2 minutes. The TRAP stain was prepared by adding the fast garnet / sodium nitrite solution to the pre-warmed deionised water along with 0.5ml naphthol AS-BI phosphate solution, 2ml acetate solution and 1ml tartrate solution, whilst mixing. 1ml of the prepared TRAP stain was then added to the fixed and washed osteoclast preparations, which were subsequently incubated at 37°C for 30-60 minutes. In the presence of TRAP, highly insoluble red-brown dye deposits are formed within cells which can be evaluated microscopically.

aVp3 staining Fixed osteoclast cultures were permeabilised in 20mM Hepes, 300mM sucrose, 50mM NaCl, 3mM

MgCl] and 0.5% Triton-xlOO for 5 minutes at 4°C. Cells were then washed four times in PBS and blocked for 30 minutes at room temperature in PBS/ 5% FCS/ 0.1% sodium azide. 250|xl mouse anti-human aVp3 (Chemicon, Temecula, CA) or IgGi isotype control (M0PC21) at 20|ig/ml were added to the wells which were incubated at room temperature for 30 minutes, before washing three

55 times with PBS. 250}xl/well of HRP-conjugated goat anti-mouse IgG (Jackson Immuno-Research) was added to each well at a dilution of 1 in 50 in PBS/ 0.5% FCS and the wells were incubated for

30 minutes at room temperature. AEC (3 -amino-9-ethylcarbazole) substrate was prepared as follows: 1 AEC tablet was dissolved in 5ml DMF (stored at room temperature and covered in foil); for use, a 4mg/ml solution was prepared in 50mM NaAC pH 5.0 plus 0.1% H 2O2. The wells were washed three times with PBS, then 250|il of AEC substrate was added to each well and the plates were incubated for 10-15 minutes at room temperature. Wells were washed twice with deionised water. Wells were then counterstained with 250|xl/well Geimsa solution (diluted 1 in 20 in deionised water and filtered through a 0.2|im filter) for approximately 30 minutes at room temperature. Wells were washed four times with deionised water and evaluated microscopically.

AEC acts as an electron donor in the presence of HRP and reduction of AEC causes a rust red p recipitate; Geimsa stained nuclei appear pale blue. Where appropriate, photographs were taken using an Olympus inverted light microscope, with phase contrast, and an Olympus SC35 camera.

2.2.11 Osteoclast pit formation

Osteoclasts were removed from 24-well plates by treatment with accutase (TCS Biologicals,

Buckingham, UK). 1ml accutase was added to each well, incubated at 37°C for 15 minutes, and cells were removed gently with a Pasteur pipette and washed in complete medium (see Section

2.2.9). Sterile dentine discs (IDS, Tyne and Wear, UK) were placed into individual wells of a 96- well tissue culture plate (1 disc/well) and pre-incubated at 37°C for 60 minutes in complete medium. Dentine discs were seeded at 2 x 10^ cells/well in complete medium containing lOOng/ml

RANKL plus 50ng/ml M-CSF and incubated at 37°C/ 5% CO 2 for 72 hours. Discs were then washed three times with PBS. For TRAP staining, discs were fixed for 20 minutes at room temperature in PBS/ 3.5% paraformaldehyde/ 2% sucrose and stained in 24-well plates as described in 2.2.9. For pit analysis, cells were removed from the discs by abrasion using filter paper; the discs

56 were stained using wheatgerm agglutinin-HRP; 100|il wheatgerm agglutinin-HRP (Vector Labs,

Burlingame, CA) at 1 in 1000 in PBS was added to each disc in a 96-well tissue culture plate and discs were incubated for 30 minutes at room temperature, washed three times with deionised water before lOOjil of AEC substrate was added (see 2.2.9) and the discs were incubated for 15 minutes at room temperature, washed with deionised water and examined microscopically. Wheatgerm agglutinin has been shown to bind to the exposed matrices that become revealed in demineralised osteoclast resorption lacunae (Selander et al., 1994) and presents as a rust red precipitate. The area of dentine resorbed was estimated using a graticule. Pictures were taken using a Zeiss axioscope with a JVC 3 chip colour camera (magnification x40).

2.2.12 RNA extraction and characterisation

For monocyte and dendritic cell samples, total cellular RNA was extracted using an RNeasy mini kit (Qiagen, Crawley, UK) following the manufacturer’s protocol. For osteoclast samples, where extraction of the RNA in situ prevented degradation, total cellular RNA was extracted using

RNAzol™B (Biogenesis Ltd, Poole, UK). Briefly, osteoclast cultures were washed twice with PBS/

0.2% EDTA, aspirated, 1ml PBS/ 0.2% EDTA was added to each well, then incubated at 37°C/ 5%

CO2 for 15 minutes. The PBS/ EDTA solution was removed from the wells and 0.5ml RNAzol™B was added to the first well, aspirated up and down with a Pasteur pipette and transferred to the second well: 0.5ml RNAzol™B was used to lyse the cells in approximately 4 wells. The lysed cells were transferred to a 1.5ml eppendorf tube and 0.1ml chloroform was added, the tube was covered tightly and agitated for 15 sec, left on ice for 15 minutes and centrifuged at 12000g and 4°C for 15 minutes. The upper aqueous phase was transferred to a fresh tube, and an equal volume of isopropanol was added. The sample was incubated at 4°C for 15 minutes, and then centrifuged at

12000g and 4°C for 15 minutes. Following aspiration the pellet was reconstituted in lOOjil DEPC water (Ambion, Austin, TX). 10|xl of 3M sodium acetate pH 5.2 and 250|ll of 95% ethanol (ice

57 cold) were added and the RNA was stored at -70°C until use. Before use, the RNA was pelleted,

air-dried and washed in 70% ethanol, then re-pelleted, air-dried and reconstituted in 100|il DEPC

water.

To check the integrity and size distribution of the total RNA, presence of distinct 18S and 28S

ribosomal RNA subunits were confirmed by denaturing-agarose gel electrophoresis and ethidium

bromide staining. A 1.2% formaldehyde agarose (FA) gel was prepared (10ml lOx FA running

buffer (Ambion) plus 1.2g agarose in 100ml RNase-free water (Ambion), microwaved to melt the

agarose, cooled to 65°C in a waterbath before 1.8ml of 37% (12.3M) formaldehyde was added and mixed thoroughly before casting in a gel support) and equilibrated in 1 x FA running buffer for 30 minutes. RNA samples were diluted in 5x RNA loading buffer (Ambion), 10|ig/ml ethidium bromide added to the RNA, which was incubated for 10 minutes at 65°C, chilled on ice and loaded

onto the equilibrated FA gel. The gel was run at 80 volts for 1-2 hours; a total RNA size marker

(Ambion) was included. The 18S and 28S ribosomal RNA species should appear as sharp bands on the stained gel with the 28S band being approximately twice the intensity of 18S RNA.

2.2.13 RT-PCR for RANK and calcitonin receptor

RT-PCR was performed using the RETROscript^'^ first strand synthesis kit for RT-PCR (Ambion,

Austin, Texas), following the manufacturer’s protocol. Briefly, l-2|xg total RNA plus 2|Ltl random

decamers were made up to 12|j,l with nuclease tree dH 2 0 , mixed, spun briefly, heated at 85°C for 3 minutes, removed to ice, spun briefly again and replaced on ice. The remaining reverse transcription

(RT) components were added: 2|il lOx RT buffer, 4fxl dNTP mix, l|Lil RNase inhibitor and Ijxl reverse transcriptase enzyme, the tube was mixed gently, incubated at 42-44°C for one hour, then

92°C for 10 minutes and transferred to -20°C. The PCR reaction was set up using 2-10|Xl of the RT reaction sample, 2 |l i 1 lOx PCR buffer (Roche-Applied-Science, Basel, Switzerland), 0 .5|l i 1 dNTP

58 mix (Roche-Applied-Science) and made up to a volume of 20|il with nuclease free dHiO, then Ijil of the appropriate PCR primers were added (5|xM of each primer) with 1-2U of thermostable DNA

Taq polymerase (Roche-Applied-Science). GC polymerase and GC PCR buffer (Clontech, San

Jose, CA) were used for the RANK reactions. For each PCR reaction a negative control, containing no reverse transcriptase, was included. The reactions for calcitonin receptor (CTR) (35 cycles) and

G3PDH (30 cycles) were cycled as follows;

Denature: 94°C for 1 minutes cycling: 94°C for 30 seconds, 60°C for 30 seconds, 72°C for 60 seconds.

Final extension: 72°C for 10 minutes

The reactions for the GC rich RANK (30 cycles) were cycled as follows;

Denature: 94°C for 1 minutes

cycling: 94°C for 30 seconds, 6 6 °C for 15 seconds, 6 8 °C for 3 minutes.

Final extension: 6 8 °C for 4 minutes

Primers for RT-PCR;

Calcitonin Receptor 5 ’ primer 5 ’-TGCGGTGGTATTATCTCTTGG-3 ’

(Sigma-Genosys Ltd) 3 ’ primer 5 ’-TTCCCTCATTTTGGTCACAAG-3 ’

RANK 5’ primer 5’-GGTGGTGTCTGTCAGGGCACG-3’

(Sigma-Genosys Ltd) 3 ’ primer 5 ’-GGCAGCCTGTGCAGACATCTG-3 ’

G3PDH (Ambion) 5’ primer 5’-ACCACAGTCCATGCCATCAC-3’

3 ’ primer 5 ’-TCCACCACCCTGTTGCTGTA-3 ’

To evaluate the RT-PCR reaction, an aliquot was analysed by horizontal gel electrophoresis.

Briefly, samples were loaded in 6 x ficoll loading dye and separated at 100 volts on a native 1%

59 (w/v) (for RANK) or 2% (w/v) (for CTR and G3PDH) agarose gel in TAB buffer (40mM Tris-

acetate, 2mM EDTA, pH7.5). l^ig/ml ethidium bromide was included in the gel to allow

visualization of DNA fragments using a 250nm UV light box. A 100 bp ladder (Invitrogen) was

included on every gel to allow estimation of amplified DNA fragment size.

2.2.14 cDNA probe synthesis, microarray hybridisation and quantification

Radiolabelled cDNA probes were synthesised in 30pl reactions containing 0.1-lpg total cellular

RNA, 2pg poly (dT) (Research Genetics), 0.5mM each dATP, dGTP and dTTP (Amersham), 50|iCi

a-*^P-dCTP (NEN, Boston, MA), 60units SUPERNaseIn (Ambion, Austin, TX), lOmM DTT, 6 |il

Sxfirst-strand buffer and 300units Superscript II (Invitrogen, Paisley, UK). The reaction was

incubated at 42°C for 2 hours before the RNA template was hydrolysed at 60°C for 30 minutes in

0.05% SDS/ 25mM EDTA/ 0.45M NaOH. The reaction was neutralised by HCl to 0.45M, with Tris

pH7.5 to 250mM and 15|ig human Cot-1 DNA added (Invitrogen). Unincorporated nucleotides

were removed with Bio-Spin 6 chromatography columns (Bio-Rad, Hercules, CA). In-house

microarrays printed with approximately 500-900 bp specific sequences corresponding to 3’-UTRs

for all known human chemokines and chemokine receptors, plus related receptors, GPCRs and their

ligands, and suitable housekeeping genes (see Appendix 2 for list and Genbank accession numbers)

were prehybridised at 55°C for 2-3 hours in 8 ml ExpressHyb solution (Clontech) with 200|ig

salmon testes DNA (Sigma, Poole, UK) and lOjig poly (dA) (Research Genetics). Radiolabelled

cDNA probes were denatured at 95°C for 3 minutes, then hybridised to the microarray for 16-20 hours at 55°C. Microarrays were washed with 0.2x SSC/ 0.1% SDS for 1 hour; twice at 55°C and

once at 65°C. Hybridised microarrays were exposed to low energy phosphorimage screens and

scanned by Storm860 (Molecular Dynamics). Data was extracted using IMAGENE software

(BioDiscovery, Los Angeles, CA) and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as the percent average of this

60 panel. Data manipulations were performed with Microsoft Excel, and hierarchical clustering achieved using GeneSpring software (Silicon Genetics, Redwood City, CA).

2.2.15 Northern blot hybridization

Northern blots were generated by electrophoresis and transfer of 5|ig of total cellular RNA, separated on a 1.2% agarose/ formaldehyde gel and transferred to Zeta-Probe GT membrane (Bio-

Rad) according to manufacturer’s instructions. *^P-labelled probes were synthesized with a Strip-EZ

DNA kit (Ambion). A 905 bp GPR105 probe consisted of nucleotides 1427-2332 from NCBI accession number XM_00309, a human G3PDH cDNA control probe (Clontech) was also used.

Probes were purified using Quickspin columns (Boehringer Mannheim), following the manufacturer’s instructions. Hybridisation was performed overnight at 65°C. Hybridised membranes were stringently washed, wrapped in plastic wrap, and then immediately exposed to X- ray film for autoradiography.

2.2.16 DC chemokine ELISAs

For chemokine assays, day 9 immature and mature (MCM-treated) DCs were washed thoroughly and re-plated in 24-well plates at 5 x 10^ cells/well with or without lOOng/ml IFN-y with or without

0.2% w/v Staphylococcus aureus Cowan’s (SAC)strain or lOOng/ml EPS. Culture supernatants were collected at 48 hours, centrifuged at 400g to obtain cell-free supernatants, stored at -20°C prior to assay by ELISA. Duoset™ ELISA development kits from R&D Systems were used for CCL2,

CCL3, CCL4, CCL13, CCL15, CCL17, CCL18 and CCL22, and R&D Systems Quantikine™

ELISA kit was used for CCL23 measurement.

61 2.2.17 Characterisation ofHEK293 cell lines

Human embryonic kidney cells (HEK293) (ATCC) and an in-house HEK293 cell line, co­ transfected with the promiscuous G protein subunit G(%i 6 and haemagglutinin (HA)-tagged GPR105 cDNA constructs, were cultured in DMEM with 10% FCS, 2mM L-glutamine and 1% non-essential amino acids. For the transfected HEK293 cell line, the selection reagents puromycin (l|Xg/ml) and geneticin (G418) (1 mg/ml) were also included in the growth medium. The presence of the HA- tagged GPR105 construct on the surface of the transfected cell line was confirmed by immunofluorescent staining and flow cytometry. An unconjugated rat anti-HA antibody (Roche-

Applied-Science, Basel, Switzerland) and a PE-conjugated anti-rat IgG antibody (Insight biotechnology) were employed. Briefly, cells were washed with ice-cold wash buffer (PBS

(Invitrogen) containing 5% FCS and 0.1% sodium azide) and then incubated with optimal concentrations of rat anti-HA tag or rat IgG isotype control for 30 minutes on ice. Cells were washed in ice-cold wash buffer and incubated with PE-conjugated anti-rat IgG for 30 minutes on ice, again washed in ice cold buffer and analysed in a FACScalibur equipped with CellQuest software. Live cells were gated on their forward and side light scatter properties.

2.2.18 Ca^^ flux measurement

The response of cells to endothelin (Tocris, Bristol, UK); UDP-glucose (Sigma-Aldrich or ICN biomedicals; detailed in appropriate Figure legends); histamine; CCL3 (R&D Systems) and CCL19

(R&D Systems) was investigated by measuring intracellular calcium (Ca^^(i)) changes using the

Ca^^-sensitive dye Fluo4-AM, in conjunction with a fiuorometric imaging plate reader (FLIPR)

(Molecular Devices, Berkshire, UK). Where indicated, cells were pre-treated with lOng/ml pertussis toxin for 18 hours. For dendritic cell and monocyte experiments, agonists were prepared in HBSS/

0.2% BSA/ ImM MgCy ImM CaCy 20mM HEPES. Fluo-4 (Molecular Probes, Oregon) was dissolved in DMSG (HYBRI-MAX™) and pluronic acid (20% v/v in DMSG). Cells were then

62 washed in the above buffer and incubated for 1 hour at 37°C/ 5% CO 2 with Fluo-4 (final concentration 4|iM) before washing twice and plating into 96-well black-walled clear-bottomed microplates (Coming-Costar, Acton, MA) at 0.5 - 1 x 10^ cells/100|il/well. The plates were centrifuged at 21 Og for 3 minutes, with no brake, and rested at 37°C/ 5% CO 2 for 10 minutes. For experiments utilising the HEK293 cell lines, 0.71g/l of the organic anion transport inhibitor probenecid was included in the buffer throughout and the cell lines were plated into 96-well black- walled clear-bottomed microplates at 2 x 10* cells/100|il/well and incubated overnight at 37°C/ 5%

CO2 . Plates were then aspirated and incubated for 1 hour at 37°C/ 5% CO 2 in 4|iM Fluo-4, washed

twice using a multichannel pipette with a final volume of 1 0 0 |il/well of buffer added to the plates and as with dendritic cells and monocytes the plates were then placed at 37°C/ 5% CO 2 for 10 minutes. For all experiments the plates were loaded into the FLIPR and a signal test was taken and laser power adjusted to obtain a basal level of approximately 1 0 , 0 0 0 fluorescence intensity units

(FIU). The cells were then excited at 488 nm using the FLIPR laser and fluorescence emission determined using a CCD camera with a bandpass interference filter (510-560 nm). Fluorescence readings were taken at 1.5 second intervals for 70 seconds and a further 10 readings were taken at 6 second intervals. Agonists or buffer (negative control) were added after 15 seconds using the

FLIPR. Max -Min fluorescence data, with negative control subtraction, was exported for each well.

Data was then imported into Microsoft Excel and EC50 concentrations were calculated using XLfit version 2.0.7a and a one site dose response model: Y - A+((B-A)/(1+((C/X)°)).

2.2.19 Preparation of necrotic cell supernatant

PBMC were prepared as described in Section 2.2.1. Cells were washed twice in HBSS (5 x 10^ cells/tube). Viable control cells were reconstituted in 5ml HBSS, while for necrotic cells the pellets were subjected to four freeze thaw cycles from -80°C to 37°C, and then reconstituted in 5ml HBSS.

Tubes were spun at 400g and supernatants were collected and stored at -80°C.

63 2.2.20 Osteoclast chemokine ELISAs

For chemokine assays, osteoclast supernatants were removed from 24-well cultures (as detailed in

Section 2.2.9), centrifuged at 400g to obtain cell-free supernatants and stored at -20°C prior to assay. Where appropriate, day 6 osteoclast cultures were stimulated with recombinant human VEGF

(R&D Systems), recombinant human IL-ip (R&D Systems) or recombinant human TNF-a (R&D

Systems) and incubated for a further 48 or 72 hours at 37°C/ 5% CO 2 before removal of the culture supernatant. Duoset™ ELISA development kits (R&D Systems) were used for CXCLl, CXCL5,

CXCL7, CXCL8 , CXCL9 and CXCLl0 quantification, following the manufacturer’s instructions.

2.2.21 Endothelial cell chemotaxis

HUVECs were fluorescently labelled as follows: 80% confluent HUVECs, in a T175 tissue culture flask, were washed in endothelial basal medium (EBM) (Cambrex, East Rutherford, NJ) and labelled in situ with 10ml of 0.5|ig/ml Calcein-AM (Molecular Probes, Eugene, OR) in EGM

(endothelial cell growth medium; consisting of EBM with addition of growth factors, cytokines and supplements (Bulletkit^^)) (Cambrex) plus 2% FCS and 0.71g/l probenecid, for 60 minutes at 37°C/

5% CO2. Cells were removed from the tissue culture flask by trypsinisation and washed HUVECs were suspended in EBM/ 0.1% BSA (fatty acid free), or EGM/ 0.1% BSA, or EGM/ 2% FCS (cell diluents) and plated at a density of 2 x lO'* cells/SOjil/well on top of the insert in a 3|im pore size 96 well fluoroblock plate (BD Biosciences). Cell diluents plus 10% FCS were employed as chemoattractants; therefore cell diluents plus 10% FCS or cell diluents were added to the bottom of the fluoroblock chambers using the sample ports. Plates were incubated for 4 hours at 37°C/ 5%

CO2. Leaving the plates intact, with the fluoroblock membrane preventing contribution to the fluorescence intensity by fluorescently labelled cells that have not migrated (cells in the top chambers), the fluorescence intensity of the bottom chamber was measured using an LJL analyst and bottom read optics at 485nm (excitation) and 530nm (emission).

64 2.2.22 Neutrophil chemotaxis

Neutrophils were purified from peripheral blood by layering on Ficoll-Hypaque, centrifuging at

400g before discarding the plasma, the PBMC layer and Ficoll, while retaining the red cell and neutrophil pellet followed by lysis of the red cells. Recombinant chemokines or osteoclast

supernatants were placed in a Costar^"^ transwell (Coming, Acton, MA) and antibodies were added

where appropriate and pre-incubated for 60 minutes at 37®C. Inserts (3|xm pore size) were placed

into the wells and 1 0 0 |il neutrophils at 1 x 1 0 ^ cells/ml were added to the top of each transwell.

Plates were incubated at 37 °C for 1 hour, with ImM EDTA added to the wells for the last 5

minutes. Inserts were removed and the cells resuspended and counted on Becton Dickinson

FACScalibur for 30-seconds each tube, ensuring that the pressure and flow rate remained constant.

2.2.23 Immunocytochemistry

Osteoclast differentiation cultures were grown on 6 -well culture slides (Falcon) for 9 days, in the

presence of GolgiPlug™ (Ijil/ml) for the last 3 hours of culture. Slides were fixed in PBS/ 3.5%

paraformaldehyde/ 2% sucrose for 20 minutes at room temperature, then permeabilised in 20mM

Hepes/ 300mM sucrose/ 50mM NaCl/ 3mM MgCy 0.5% Triton-xlOO for 5 minutes at 4°C.

Samples were blocked in PBS/ 5% FCS/ 0.1% sodium azide for 1 hour at room temperature, before

incubation with either 10|ig/ml goat anti-human CXCL5 (R&D Systems) or goat anti-human

CXCL8 (R&D Systems) or goat IgG control for 1 hour at room temperature. Bound antibody was

revealed using HRP-conjugated mouse anti-goat IgG (Jackson Immuno-Research) and AEC

substrate (see Section 2.2.10). Slides were counterstained with Haematoxylin Gill no. 3, rinsed in

tap water and mounted in aqueous feramount (Dako, Glostmp, Denmark).

65 Chapter 3 in v/tro-derived osteoclasts and dendritic cells

66 Chapter 3

3.1 Introduction

The recent advances in developing in vitro differentiation systems to mimic fully functional specialised cell types, reveals the plasticity of peripheral blood monocytes (PBM). This cell type is capable of in vitro differentiation into macrophages, myeloid dendritic cells, osteoclasts, microglial cells, foam cells and macrophage giant cells (Palucka et al., 1998; Nicholson et al., 2000; Kaur et al., 2001; Hayden et al., 2002; Dugast et al., 1997). Two of the most functionally diverse of these cell types are the dendritic cell and the osteoclast. These cell types exist in distinct tissue locations and isolating large numbers of purified cells from human and rodent tissue, while possible, is difficult. In addition, they exist in vivo in multiple stages of differentiation / maturation and are therefore difficult to study due to heterogeneity. The development of techniques to differentiate these cell types in vitro from PBM has greatly advanced the understanding of these cell types and their activation requirements. Indeed dendritic cells differentiated from peripheral blood monocytes demonstrate in vivo functionality, as represented by efforts to use in vfrm-derived antigen pulsed dendritic cells as immunotherapy’s for cancer or to modulate the immune response (Brossait, 2002).

Dendritic cells have long been described as sentinels of the immune system (Banchereau and

Steinman, 1998). They function as antigen presenting cells, sampling the microenvironment for foreign antigens, migrating to lymph nodes and presenting processed peptide / MHC complexes to

T cells. While other cell types may present antigen to T cells, the dendritic cells are far more potent and possess the exceptional ability to present to unprimed, naïve T cells. This ability to both sample antigen at the site of infection and inflammation and present to naïve T cells in the lymph node makes these cells unique and is a result of maturation in response to inflammatory signals (Sallusto and Lanzavecchia, 1994). The differentiation of myeloid dendritic cells from PBM is now a well

67 characterised in vitro technique. Immature DCs can be derived in vitro from CD 14-positive PBM in the presence of GM-CSF and IL-4 (Sallusto and Lanzavecchia, 1994). Immature DCs can then be matured or activated by many different factors to initiate a change in phenotypic cell-surface molecules and produce a functionally discrete cell type. The term immature is used for DCs characterised morphologically by their unique dendritic appearance, by their low expression of

CD80 and CD 8 6 co-stimulation molecules, intermediate expression of MHC Class II (HLA-DR) and their efficiency at antigen uptake. These cells are of comparable phenotype and functional similarity to the DCs that exist in the blood and tissues, such as Langerhans cells of the skin

(Austyn, 1998). Mature or activated dendritic cells show increased expression of the co-stimulatory molecules CD80 and CD 8 6 (B7.1, and B7.2, respectively) and express high levels of CD83. These cells show a reduced ability to endocytose antigen and an increased ability to present peptide /

MHC complexes to naïve T cells. These cells can be compared to the interdigitating DCs present in the T cell zones of the lymph nodes, capable of interacting with naïve T cells. Whilst the terms immature and mature have been traditionally used for these cell types, and evidence exists that the mature can be induced from the immature population (Reddy et al., 1997; Bender et al., 1996; Celia et al., 1997), they may also represent functionally distinct states of activation and relevant terminology such as activated or primed could be used to describe mature DCs.

Osteoclasts are specialised multinucleated giant cells that adhere to bone matrix. There are many defining characteristics of osteoclasts, such as their multinuclear nature, expression of tartrate resistant acid phosphatase (TRAP), calcitonin receptors (CTR) and vitronectin receptors (aVP3)

(Suda et al., 1992; Horton et al., 1993). Perhaps the most fundamental characteristic of the osteoclast is its ability to resorb bone. Recently, systems for differentiating osteoclasts from PBM have been developed. Multinucleated cells (MNC) expressing TRAP, CTR and aVp3, and capable of bone resorption can be differentiated from human PBM in the presence of M-CSF and RANKL

(Quinn et al., 1998; Flanagan and Massey, 2003). Although other cytokines, such as TNF-a and IL-

68 ip have been shown to enhance osteoclast differentiation in vitro, RANKL is necessary for the generation of MNC from their monocyte precursors (Fuller et al., 1998). Where factors such as

TNF-a have been used to drive this differentiation process, it is reasonable to postulate that soluble

RANKL has also been present, or induced as a result of a contaminating stromal cell or T cell population (Horwood et al., 1999). These in vitro differentiated cells are, for the most part, phenotypically and functionally comparable to the multinucleated osteoclasts found at the site of bone turnover and have been used extensively to further understand this specialised cell type.

This chapter aims to present experimental procedures for the differentiation of osteoclasts and dendritic cells from their monocyte precursors, and the subsequent characterisation of these polarised cell populations. Whilst published protocols are being followed, a brief presentation of phenotypic and functional data demonstrates that the in vitro differentiated cells are representative of those in the scientific literature.

3.2 Results

3.2.1 Purification o f monocytes from peripheral blood

Monocytes were enriched from PBMC using a MACS™ magnetic bead system and negative selection kit. Analysis of the starting PBMC population and the purified monocyte population by flow cytometry is shown in Figure 3.1. Forward scatter (FSC) / side scatter (SSC) profiles demonstrate the enrichment of a population with higher forward and side scatter properties, indicators of size and granularity, respectively (Fig. 3.1 A, B). These enriched cells can be identified as CD 14-positive monocytes by immunofluorescent staining (Fig. 3.1 C-F).

Contaminating cells are mostly CD45-intermediate, CD 14-low, CD3-negative and CD 19-negative granulocytes. Monocyte preparations used throughout this study yielded 80 - 95% CD 14-positive monocytes (data not shown).

69 PBMCs Monocytes

Data.003 Data.003 B.

5g-

100 150 200 250 50 100 150 200 250 FSC-Height FSC-Height

Data.003 Data.004 c. 19% D.

P É É B L 50% 10' 10^ 10" FL1-CD3 FL1-CD3

Dala.007 Data.003 E, F.

8 6 % ^ 2

FL1-CD45

Figure 3.1 Two colour flow cytometric analysis of PBMC and purified monocytes Monocytes were purified from PBMC using the MACS magnetic bead system and negative isolation kit. The starting PBMC population and the purified monocyte population were subsequently analysed by flow cytometry. FSC / SSC profiles of (A) PBMC and (B) purified monocytes are shown. B cell (CD 19 positive) and T cell (CD3 positive) populations are demonstrated in the (C) PBMC and (D) purified monocyte fraction. Monocyte populations (CD45- positive, CD 14 positive) are shown for (B) PBMC and (F) purified monocyte fraction. Percentage values for each population are shown in the relevant quadrant. Data shown is one representative of numerous donors.

70 3.2.2 Differentiation and characterisation of dendritic cells from peripheral blood

monocytes

Myeloid dendritic cells were differentiated from MACS™ purified monocytes (Fig 3.1) in the presence of GM-CSF and IL-4 for 6 days, replenished with fresh cytokines on day 3. Day 6 DCs were transferred to fresh culture wells with fresh cytokines and cultured for a further 3 days in either the presence (mature) or absence (immature) of monocyte-conditioned medium (MCM).

These dendritic cell preparations demonstrated the classic stellite dendritic cell morphology when examined by light microscopy. They were characterised by immunofluorescent staining for CD83,

CD80, CD 8 6 and HLA-DR expression. Immature DCs are CD83-negative, CD 8 6 -low, CD80-low, and HLA-DR-positive, whereas mature DCs are CD83-positive, CD80-positive, CD 8 6 -positive and

HLA-DR-high (Fig. 3.2). The maturation state of these two DC populations can be characterised functionally by their ability to internalise and present processed antigen (Celia et al., 1997;

Banchereau and Steinman, 1998; Rescigno et al., 1997). The FITC dextran uptake assay is characteristically used as an indication of the antigen uptake capabilities of these cell types.

Immature DCs are more efficient at endocytosing antigen than their mature counterpart; the DCs differentiated in this study follow this trend (Fig. 3.3). In humans, a convenient measure of a naïve

T cells response to antigen is the proliferative response to allogeneic cells. Mature DC preparations derived using this in vitro differentiation system induce a greater proliferative response in this assay than their immature counterparts (Fig. 3.4); this is indicative of their enhanced antigen presentation capabilities. Mature DCs also stimulate more IFN-y production from naïve T cells than immature

DCs (Fig. 3.5); a readout of functionally stimulated T cell responses.

A further benchmark of functionally complete DCs is the production of IL-12. MCM treatment increasingly stimulates the production of IL-12 p40 (Fig. 3.6), however this biologically inactive subunit is not known to confer Thl polarising capabilities to the DC (Spisek et al., 2001). The

71 bioactive counterpart, IL-12 p70 is not produced at measurable levels in these cultures (Fig. 3.7).

However, these in v/fro-derived immature DCs are capable of producing IL-12 p70 on stimulation with IFN-y and Staphylococcus aureus Cowan’s (SAC) strain (Fig. 3.7).

72 A. IgGl PE CD83 PE 3.4 W O 2 Immature DC .Nmi ii'ii,., II 10'-' 10 ’ 10*^ 10'^ 10" ' 10" 1 0 ' 10^ 10" 10

Mature DC

10"4 r,U in'4 n I 4 in'- 4ri*^ in" 4 io^r 4 10So in' in'- in" io ISLog fluorescence Log fluorescence

IgG2a PE HLA-DR PE B.

Immature DC

10" 1 0 ' 10^ 10" 10^ 10" 1 0 ' 10^ 10" 10

4.3 Mature DC

10^^1 P " ly ™ l0^ 1 o'^ IqO 1 0 ^ io 2 io3 io"l Log fluorescence Log fluorescence

IgG I FITC CD80 FITC CD8 6 FITC

j 3.6 j 25.4 26.2 Immature DC

'0 '1 '2 '3 4 '0 S 2 s o 0 ■ Tnl' ■ ■ "TJT 10"4 10' 10^ 10" 10^ 10" i,J 10' i IO'-... 10" 10^ 10" 10' 10"^ 10" 10 J 3.6 J 51,8 Mature DC

y U -4 SO j X '1 S 2 SO - 4 '0 '1 '2 SO 10" 1 0 ' 10^ 10" 10^ 10" 1 0 ' 10'^ 10" 10^ 10" 1 0 ' 10^ 10" 10 Log fluorescence Log fluorescence Log fluorescence

Figure 3.2 Cell surface phenotypes of immature and mature DCs

DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 6 Immature DCs were incubated with (mature) or without (immature) MCM for 72 hours and FACS analysed for (A) CD83, (B) HLA-DR and (Q CD80 and CD 8 6 expression. Isotype control antibody profiles are shown in the left panel: Mean fluorescence intensity (MFI) is indicated in the top right hand comer of each histogram. Data shown is representative of many donors.

73 100

Ü_

immature mature

Figure 3.3 Comparison of antigen uptake capabilities of immature and mature DCs

DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 6 DCs were incubated with (mature) or without (immature) MCM for 72 hours. Immature and mature DCs were then incubated with FITC dextran for 30 minutes at 37°C, washed and analysed for internalised FITC-dextran by flow cytometry. MFI data is shown, for the mean + SEM of 3 experiments.

150000-1 — imm DCs +T cells

m 125000- —^ mature DCs +T cells

100000 - —o - immature DCs alone

r g 75000- ■ - A - mature DCs alone P g- 50000- • T cells alone

25000- □ Medium alone

number of DCs / well

Figure 3.4 Proliferation of naïve T cells in response to allogeneic immature or mature DCs Untouched T cells were purified using the MACS magnetic bead system and a pan T cell kit (depletion of B cells, NK cells, DCs, monocytes and granulocytes). DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 9 immature (imm) or mature (MCM treated) DCs were washed extensively, irradiated (‘^^Cs) and incubated with allogeneic T cells in round bottomed 96-well plates for 5 days. Proliferation was measured by incorporation of thymidine over the final 16 hours. Data shown are mean +/- SEM of 3 experiments.

74 1 .00-1

0.75 O) c 0.504

0.25

0.00 immature DCs mature DCs

Figure 3.5 IFN-y production from T cells in response to allogeneic immature or mature DCs Untouched T cells were purified using the MACS magnetic bead system and a pan T cell kit (depletion of B cells, NK cells, DCs, monocytes and granulocytes). DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 9 immature or mature (MCM treated) DCs were washed extensively, irradiated ('^^Cs) and incubated with allogeneic T cells, at a ratio of 1 DC to 100 T cells, in round bottomed 96-well plates for 5 days. Supernatant was removed on day 5 and assayed for IFN-y levels by ELISA. Data shown are mean of two independent experiments. The error bars represent minimum and maximum values.

1 0 . 0-1 I § 5.0- Q. CNJ ^ 2.5-

0.0 0 24 48 72

Time post MCM addition (hours)

Figure 3.6 IL-12 p40 production from DCs following maturation with monocyte-conditioned medium

DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with MCM and supernatant was removed at various time points. Samples were assayed for IL-12 p40 by ELISA. Data are mean +/- SEM of three experiments.

75 600n

5 400 oN Q. CN T 2004

Figure 3.7 Induction of IL-12 p70 from monocyte-derived DCs

DCs were differentiated from purified PBM by incubation with GM-CSF and IL-4. Day 6 immature DCs were stimulated with 100 ng/ml IFN-y and / or 0.2% w/v SAC. Supernatants were removed at 48 hour and assayed for IL-12 p70. Data are mean + SEM of three experiments.

76 3.2.3 Differentiation and characterisation of osteoclasts from peripheral blood monocytes

Osteoclasts were differentiated in vitro from MACS™ negatively isolated PBM, in the presence of

M-CSF and RANKL, in tissue culture treated plates. These in vfrro-derived osteoclast preparations were routinely characterised for MNC, aVp3 expression and tatrate resistant acid phosphatase

(TRAP) expression. The cultures contain predominantly TRAP-positive, aVpS-positive MNC; with some contaminant of mononuclear cells (Fig. 3.8). Day 9 is optimal for multinuclear cell formation with greater than 70% of the nuclei in the culture contained within MNC (Fig. 3.9). At day 10 a decrease in the number of TRAP-positive MNC / field reflects the reduced viability of these cells and detachment of dead cells during the staining procedure. To assess the resorption capability of these in v/fro-derived osteoclasts, the cells were removed from the tissue culture plastic and re­ seeded onto dentine slices. After a further 72 hours of incubation, TRAP-positive MNC and resorption lacunae could be visualised (Fig. 3.10): the shaded striations seen in figure 3.10 reflect the inevitable variation in thickness across the dentine slice following cutting. The optimal time point for removal of the osteoclasts from tissue culture plastic, considering TRAP-positive MNC formation, resorption capabilities and viability, was day 9 (Fig. 3.11). Extracted RNA from these osteoclast preparations was shown, by RT-PCR, to contain mRNA species encoding CTR and increased levels of RANK mRNA, the receptor for RANKL. RNA extracted from day 0 monocyte populations did not contain detectable CTR or RANK mRNA transcripts (Fig. 3.12).

77 Control Ig anti-aV(33

. ^ B. 1 / » \

y f 1

/ •,-*■ i t ^ %

^ 1

1 V..

V.--

TRAP staining c.

C * it % v W (5 0

!■ =« \ ■* 4 A ^ #

Figure 3.8 Staining of M-CSF-plus RANKL-derived osteoclast preparations for aV|33- positive and TRAP-positive MNC Osteoclasts were derived from purified PBM by incubation with M-CSF and RANKL. Osteoclast preparations in 24-well tissue culture plates were fixed and either directly stained for TRAP (Q or permeabilised and incubated with irrelevant antibody (A) or anti-aV|33 (B). Bound antibody was revealed using HRP conjugated goat anti-mouse IgG and AEC substrate; samples (A) and (B) were counterstained with Geimsa to reveal nuclei. One representative experiment of many is shown.

78 -TRAP +ve MNC % Nuclei in MNC

o 300-, r100 ■a it:0

Ü 2 0 0 -

> -50

0 100 - E O 3

6 7 8 95 10 11

Day of culture

Figure 3.9 TRAP-positive multinucleated cell formation in osteoclast differentiation cultures Osteoclasts were derived from purified PBM by incubation with M-CSF and RANKL. Osteoclast cultures in 96 well tissue culture plates were fixed and stained for tartrate resistant acid phosphatase (TRAP) or nuclei using Giemsa stain. The number of TRAP-positive multinucleated cells and the percentage of nuclei present within multinucleated cells were counted using an inverted microscope at a magnification of x200. Data shown are mean +/- SEM of 3 donors.

79 A.

■m.

\ \:

B.

Resorption A:% Lacunae

Figure 3.10 TRAP-positive multinuclear cells and resorption lacunae on dentine slices Osteoclasts were derived from purified PBM by incubation with M-CSF and RANKL. Osteoclast preparations in 24-well tissue culture plates were washed and treated with accutase to detach the cells, which were then seeded onto dentine and incubated for a further 72 hours. In {A) dentine discs were fixed and stained for TRAP. In {B) cells were removed from the discs by rubbing with filter paper; the discs were then stained for resorption lacunae using wheatgerm agglutinin HRP, revealed with AEC substrate. A resorption lacunae is indicated with an arrow. Pictures were taken using a Zeiss axioscope with a JVC 3 chip colour camera (magnification x40). One example of ten is shown.

80 A. TRAP +ve MNC formation

200 U 0 Z u ^ « 0 0 + I 0 1 0 0 - g T3

O 0 Z O

9 10 time (days)

B. 114U.] don or 1 don or 2 Resorption of dentine slices

20n

8 9 10 time (days)

Figure 3.11 Time course for formation of TRAP-positive multinuclear cells and resorption lacunae on dentine slices Osteoclasts were derived from purified PBM by incubation with M-CSF and RANKL. Osteoclast preparations in 24-well tissue culture plates were washed and treated with accutase to detach the cells which were then seeded onto dentine and incubated for a further 72 hours. (A) Dentine discs were fixed and stained for TRAP and TRAP-positive multinucleated cells were counted using an inverted microscope at a magnification of x 200. (B) Cells were removed from the discs by rubbing on filter paper, the discs were then stained for resorption lacunae using wheatgerm agglutinin HRP and revealed with AEC substrate; the area of pits was quantified using a graticule. Data shown are mean + SD of 5 replicates.

81 600 A. 500 400 300 CTR 200

100 1. Monocyte + RT 2 . Monocyte no RT 600 B. 500 3. Osteoclast + RT donorl 400 4. Osteoclast no RT donor 1 300 5. Osteoclast + RT donor 2 G3PDH 200 6 . Osteoclast no RT donor 2 7. Osteoclast + RT donor 3 100 8 . Osteoclast no RT donor 3 9. Osteoclast + RT donor 4 10. Osteoclast no RT donor 4 C. 700 600 1111.. No Template 500 400 RANK 300 200 100 ttttttîtîtî 1 2 3 4 5 6 7 8 9 10 11

Figure 3.12 RT-PCR analysis of CTR, G3PDH and RANK transcripts, mRNA expression in monocytes and osteoclasts Monocytes were purified from PBMC using the MACS magnetic bead system and negative isolation kit. Osteoclasts were derived from purified PBM by incubation with M-CSF and RANKL. Total RNA was extracted from monocytes and osteoclasts using RNAzol™. A cDNA template was then synthesized using RETROscript^'^ reverse transcriptase (RT) and PCR was performed using oligonucleotide pairs specific for CTR (A), G3PDH (B) and RANK (CJ. Controls were PCR products from reverse transcription reactions with no RT. PCR products were analysed on an agarose gel containing ethidium bromide. One representative example of three monocyte preparations and four representative examples (donor 1 - donor 4) of seven osteoclast preparations tested are shown.

82 3.3 Discussion

Differentiation of osteoclasts and myeloid dendritic cells from PBM provided a readily available

source of cells for study. These in vitro differentiation systems are well documented as

representative of their primary cell type found in vivo. Fully characterised myeloid DC and

osteoclast differentiation systems have been presented.

Day 9 immature and mature (MCM treated) DCs were characterised for surface marker expression

by immunofluorescence and flow cytometry. Immature DCs were HLA-DR (MHC class 11)

medium, CD80-low, CD 8 6 -negative and CD83-negative; conversely mature DCs were HLA-DR high, CD80-high, CD 8 6 -high and CD83-positive. The two DC populations were also tested in

functional assays to assess their capability to take up antigen (FITC dextran uptake) and stimulate naïve T cell proliferation (allogeneic T cell proliferation). Mature DCs demonstrated a reduced

capacity to take up FITC dextran and an increased capacity to activate naïve T cells, demonstrated by proliferation and IFN-y production in response to alloantigen. The change in expression of cell surface molecules described and altered abilities to take up antigen and stimulate naïve T cells, correspond with a process commonly referred to as maturation (Banchereau and Steinman, 1998)

(Reseigno et al., 1997; Sallusto and Lanzavecchia, 1994). These in vfrro-derived cells represent immature and mature DC populations suitable for further investigation. There are many documented protocols for maturing DC populations. Indeed the most commonly used now, is a cytokine cocktail containing prostaglandin E 2, TNFa, IL-1(3 and IL - 6 (Choi et al., 2000). However, at the onset of this project, the most reliable method was the use of MCM (Reddy et al., 1997). A further marker for functionally active DCs is the production of the bioactive IL-12 p70 subunit.

Originally this was thought to be a marker for fully functional mature DCs (Spisek et al., 2001).

However, literature is now emerging to suggest that bioactive IL-12 is only produced in response to

83 certain stimuli, such as SAC, at early stages of maturation (Ebner et al, 2001), as was suggested by

my data (Fig.3.7).

Day 9 in vfrro-derived osteoclast preparations were characterised for TRAP expression, aV(33

expression, presence of mRNA encoding the osteoclast marker CTR and formation of MNC and

bone resorption capabilities. Osteoclast cultures contained numerous multinucleated cells

expressing both TRAP and aVP3; however these cultures were not homogeneous in nature and

contained many mononuclear cells of macrophage appearance, which were not characterised. These

osteoclasts also contained mRNA transcripts encoding CTR, which were not found in the precursor

monocyte population. The TRAP-positive osteoclasts are also capable of adhering to and resorbing bone. This corresponds to a series of characteristics and functional capabilities fully convergent

with a description of these multinucleated cells as in vffro-derived osteoclast cultures. These

osteoclast populations were therefore suitable for further investigation (Quinn et al., 1998; Flanagan

and Massey, 2003). Day 9 in vfrro-derived osteoclast preparations were also tested for the presence

of mRNA encoding RANK. These osteoclasts contained mRNA transcripts encoding RANK, which were not found in the precursor monocyte population. Little is known about the regulatory mechanism of RANK expression and although RANK is required at some point in an osteoclast precursor population, to allow response to RANKL, early stage precursor populations in the mouse

are RANK-negative. There is evidence to suggest that M-CSF induces RANK expression on osteoclast precursors as they commit to the osteoclast lineage (Arai et al., 1999). However, investigators from Amgen have characterised CD 14-positive peripheral blood monocytes, capable of differentiating into osteoclasts in the presence of M-CSF and RANKL, as positive for the

RANKL receptor RANK (Shalhoub et al., 2000). With this in mind, human peripheral blood monocytes were a poor choice of negative control for the RANK RT-PCR experiments. However, the information required to characterise the osteoclast RNA preparations was the presence of mRNA transcripts encoding RANK in these samples. Indeed the inability to demonstrate RANK

84 transcripts in the three monocyte samples tested suggests an increase of RANK message in the osteoelast samples.

The following chapter uses these in vitro differentiated cells to investigate differential gene expression in these lineage related, but functionally divergent myeloid cell types.

85 Chapter 4

Microarray profiling of monocytes, osteoclasts and dendritic cells

86 Chapter 4

4.1 Introduction

The previous chapter demonstrates how two phenotypically distinct cell types (and subpopulations thereof) may be differentiated from peripheral blood monocytes. But what genes play a role in the functions of these distinct cell types? To help address this question microarray gene technology was used to analyse the expression of a large family of soluble faetors and their reeeptors and identify differentially expressed genes in these in v//ra-derived specialised cell types.

Nucleic acid arrays work by hybridisation of labelled DNA or RNA to complementary gene fragments or expressed sequence tags (ESTs) attached to a surface. Commonly, arrays are designed based on precise sequence information and carefully selected ESTs are deposited onto specific locations on a surface such as glass or nitrocellulose. These arrays are then probed with labelled eDNA and hybridisation of complimentary DNA molecules ensues (Loekhart and Winzeler, 2000).

Using RNA purified from cellular sources as a template for cDNA synthesis allows monitoring of gene expression at particular timepoints. While not measuring the actual transcription of genes, the accumulation of cytoplasmic mRNA is a commonly used method to assess gene expression, which accounts for mRNA stability and accessibility to translation to functional protein. The gene expression profile of a cell is a dynamic entity, changing rapidly in response to stimuli or during normal cellular events, such as adhesion, migration, proliferation and even cell death. Knowing when and where a particular gene is expressed can give us clues to the biological function of its translated protein. It is therefore possible to design experiments where the significance of differential gene expression might give greater clues to the function of the encoded protein. Indeed,

DNA microarrays have been used to demonstrate differential gene expression in different tumour cell types, indicating proteins involved in tumour progression (Alon et al., 1999; Pérou et al., 1999;

87 Ross et al., 2000). In this chapter, a DNA microarray was used to identify differences in expression

of a discrete selection of genes, in myeloid lineage cells differentiated from a common precursor,

the peripheral blood monocyte. Figure 4.1 outlines the prineiples of the particular methodology

used.

The custom-made microarray I used consists of 162 known genes including all known chemokine

and chemokine receptor genes, related G protein-coupled seven transmembrane receptors, their

ligands, and a panel of 40 housekeeping genes. Expected results based on data from published

reports act as an effective internal control, providing a certain amount of validation, therefore giving

a degree of confidence in the relevance of new information obtained. With this in mind, the change

in expression of some of the chemokine and chemokine receptor genes in dendritic cells following

activation / maturation was compared to information in the scientific literature and extensive study

of produetion / expression of these chemokines and their receptors at the protein level. This study

also aimed to identify chemokines and chemokine receptors that are differentially regulated in

osteoclast cells and, therefore, may be of interest in terms of their unique biology. Osteoclasts are

related to myeloid dendritic cells, at least by way of their shared origins. However, unlike myeloid

dendritie cells, osteoclasts have so far been sparsely studied with respect to their chemokine and

chemokine receptor expression profiles. Other seven transmembrane receptors of interest and their

known ligands were included in the array to identify differential expression of these structurally

related receptors in monocytes, osteoclasts, immature and mature dendritic cell populations. This

chapter aims to analyse the differences in expression of a diserete selection of genes in these lineage related but functionally discrete cell types, as measured by microarray profiling. monocytes ‘osteoclasts’

7.

6 .

## # #

Figure 4.1 Schematic of a differential gene expression hybridisation experiment 1. The rationale behind differential gene expression experiments is to compare gene expression in two or more different cell types, such as osteoclasts and monocytes, or a single cell type following different stimuli. 2 . Total RNA is extracted from the cell preparations as described in Materials and Methods, Chapter 2. 3. mRNA is reversed transcribed to P-labelled cDNA using a-' P-dCTP and RT. 4. The cDNA probe is then hybridised to the microarrays (500 - 900 bp specific 3’ UTR sequences printed onto Hybond N+ membranes using an Affymetrix microarrayer). Hybridised microarrays are washed thoroughly and exposed to low energy phosphorimage screens and scanned hy a STORM860 phosphorimager. 6 . An example of a scanned image. Data is extracted using Imagene software. 7. Normalised data is analysed using the GeneSpring software, an example of hierarchical gene clustering is shown.

89 4.2 Results

4.2.1 Characterisation of monocyte, DC and osteoclast samples for RNA extraction

Monocyte, osteoclast and dendritic cell preparations were characterised and RNA was extracted for probing the microarray filters. Monocytes were purified fi*om PBMC by negative selection using the

MACS monocyte kit; the samples used for array profiling were characterised by FACS analysis for

CD14 expression as described in Section 3.2. Resulting monocyte preparations were 87.5% +/- 1.5

(SEM) CD 14-positive, n= 6 . Figure 4.2 shows summary data for the dendritic cell preparations subjected to microarray analysis. The immature dendritic cell preparations exhibited very low CD 8 6 expression, were negative for CD83 (that is, not significantly above background) and positive for

HLA-DR. The mature dendritic cell preparations showed markedly increased levels of CD 8 6 and

HLA-DR expression and were positive for CD83. Data for the osteoclast preparations subjected to microarray profiling is shown in Figure 4.3. A median of 61% of the nuclei present in the cell preparations could be attributed to the multinucleated cells. The uncharacterised contaminating mononuclear cells were of macrophage appearance. Of the multinucleated cells present in these preparations, over 90% were TRAP-positive. All osteoclast preparations subjected to microarray profiling demonstrated expression of the calcitonin receptor and RANK mRNA transcripts by RT-

PCR, see Chapter 3, Figure 3.12, as would be predicted for osteoclasts.

Total RNA was purified from the cell preparations using either RNeasy™ or RNAzol™ extraction methods. RNAzol™ extraction was required to obtain reasonable yields of RNA from the osteoclast preparations. Formaldehyde agarose gel electrophoresis and ethidium bromide staining were used to check the integrity and size distribution of the total RNA preparations, an example of which is shown in Figure 4.4. An estimate of the concentration of RNA present in these samples was also made from these analytical gels.

90 A. B. CD86 CD83

200n 50-1

100

immature mature immature mature

C. HLA-DR C.

400 LL

200

immature mature

Figure 4.2 Characterisation of dendritic cell preparations for gene array profiling

DCs were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 Immature DCs were incubated with (mature) or without (immature) MCM for 72 hours and analysed by immunofluorescent staining and flow cytometry for expression of (A.) CD8 6 , (B.) CD83 and (C.) HLA-DR. The data presented are mean fluorescence intensity (MPI) minus the MPI for the isotype control. Data are mean + SEM of 6 donors.

91 TRAP +ve MNC

100n Ü 75- f I 50 25-

B. Nuclei in MNC

100-1 c •O (D c 75-

1 1 50- 0 Ü 3 C 25-

0-

Figure 4.3 Characterisation of osteoclast preparations for gene array profiling Osteoclasts were derived from monocytes by incubation with M-CSF and RANKL for 9 days. Osteoclast preparations in 24-well tissue culture plates were fixed and either directly stained for TRAP (A), or permeabilised and stained with Geimsa to visualise nuclei (B). Individual preparations are represented by individual data points; the median value is indicated.

92 28 S rRNA

18SrRNA

o o 2 O > % 5 o % B (D ft 3 ft o o o oO o ’ pT 3 ft CA O CL CL CL o o o o B•"t B 1 B z:3 to U) W >

Figure 4.4 The integrity of RNA preparations subjected to microarray profiling, as examined by formaldehyde agarose gel electrophoresis RNA samples were diluted in RNA loading buffer containing lOpg/ml ethidium bromide and heated at 65°C for 3 - 5 minutes, chilled on ice and then loaded onto an equilibrated 1.2% formaldehyde / agarose gel. Gels were run at 80 volts for approximately 2 hours and photographed under short-wave UV illumination. Four representative samples are shown, but the integrity of all 28 RNA samples used in this array was examined.

93 4.2.2 Gene array profiling of dendritic cell, osteoclast and monocyte populations

Through use of a custom made gene microarray, the expression by monocytes, immature DCs,

mature DCs and osteoclasts of the entire superfamily of chemokines and their receptors, as well as

selected GPCRs was investigated.

Total RNA was extracted from these in vitro differentiated cell types and the mRNA fraction was reverse transcribed to produce a radiolabelled cDNA probe which was hybridised to the custom made microarrays, consisting of 162 specific 3’-UTR sequences (ranging from 500 - 900 base pairs) printed onto Hybond-N+ membranes using an Affymetrix microarrayer. Hybridised microarrays were washed thoroughly and exposed to low energy phosphorimage screens and

scanned by a STORM860 phosphorimager. Data was extracted using Imagene software. The resulting gene expression data is listed in tabular form (Appendix 1) and was analysed by hierarchical clustering, performed blind with respect to description of cell type. Data were grouped together solely on the similarity between gene expression profiles. This analysis resulted in the formation of four clearly defined clusters, correlating with the cell populations of immature DCs, mature DCs, monocytes and osteoclasts (Fig. 4.5). This grouping of populations into distinct clusters according to phenotype suggests significant differences in the pattern of mRNA transcripts expressed in these four cell types.

94 Osteoclast 1 Osteoclast 2 Osteoclast 3 Osteoclast 4 Osteoclast 5

Osteoclast 6 Osteoclast 7 Monocyte 1 Monocyte 2 Monocyte 5 Monocyte 4

Monocyte 6 Monocyte 3 iDC 1 day 9 iDC 2 day 9 iDC 3 day 9

iDC 4 day 6 iDC 4 day 9

iDC 5 day 6 iDC 5 day 9

iDC 6 day 9

iDC 6 day 6 mature DC 1 mature DC 2 mature DC 5 mature DC 3 mature DC 4

mature DC 6

Figure 4.5 Hierarchical clustering of mRNA transcript expression data acquired by gene microarray profiling Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs (iDC), respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Hierarchical clustering of microarray gene expression data from monocytes, osteoclasts, immature DCs and mature DCs is shown. Genes represented on the array are clustered along the horizontal axis and samples are clustered on the vertical axis of the dendogram. The bar on the right hand side gives an indication of the level of expression: blue=low, red=high.

95 4.2.3 Differentially expressed genes

From the microarray analyses, genes demonstrating clear differential expression, within the four

cell types examined, are listed in Table 4.1. These were selected as mRNA transcripts which

showed at least a 3 fold differential expression in a particular cell type when compared to all other

cell types. For instance, the immature dendritic cell expression of the GPR105 mRNA transcript is

greater than three times the expression level of this transcript in mature dendritic cells, monocytes

or osteoclasts. For these analyses, day 6 immature DCs were not included, as they do not actually represent a unique cell type. In all cases, day six immature DCs gave transcript expression levels very similar to day 9 immature DC: see Appendix 1. Genes with very low expression (less than 20

arbitrary units for each of the cell types) were excluded from Table 4.1. Only 20 of the 162 genes

evaluated by array profiling were highlighted by this stringent analysis. Many of the expected genes were illustrated here, including differential expression of chemokine and chemokine receptor mRNA transcripts in the mature and immature dendritic cell populations: see Section 4.2.4.

Other differentially expressed transcripts included the anaphylotoxin receptors C3aR and C5aR, which appeared to be downregulated in the mature DC population and H963, an orphan seven transmembrane GPCR which was upregulated on DC maturation. Of note also are two genes that

were higher in the monocyte population, the chemokine receptors CX 3CRI and hm74. However, the

transcript levels for these receptors were low and their significance would have to be further

evaluated. Of most interest in the dendritic cell populations was the recently de-orphaned G protein-

coupled receptor gene, GPR105, which showed a large increase in transcript expression in the

immature DC population when compared to the monocyte precursor population. This increase in

GPR105 expression was subsequently reversed in the mature DC population. The CXC chemokine,

CXCL5 (ENA-78) showed a very significant increase in transcript levels in the osteoclast

population; whereas CXCL 8 (lL-8 ) gene expression, while evident in osteoclasts, was highest in the

96 monocyte population. Crucially none of the 40 housekeeping genes demonstrated differential expression between the four cell types evaluated.

im m atu re m atu re m onocyte osteoclast DC DC H963 8 rv 1 3 GPR105 1 228 43 18 14 C5aR 35 31 37 C3aR 33 21 34 CCL2 33 '326 0 11 CCL3 38 5 261 48 SCYA3L 148 13 268 56 CCL1 30 0 189 19 CCL13 1 231 33 3 4 CCL15 1 82 15 0 0 CCL17 112 488 7 4 e c u s 447 58 15 21 CCL22 48 ' -343 j 11 13 CCL23 1 124 21 1 2 CCR1 53 38 55 CCR7 11 13 12 CX3CRI 2 2 25 5 hm74 9 4 25 5 CXCL5 1 6 0 46 CXCL8 0 7 175 23

Table 4.1 Differentially expressed genes in myeloid cells Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and lL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Genes demonstrating greater than a 3 fold difference of expression in one cell type, when compared to all others, are highlighted. Data are arbitrary hybridisation units, normalised to the average of the housekeeping panel and expressed as percentage average of this panel. Data are means of 6 (monocyte, immature DC, mature DC) or 7 (osteoclast) experiments. A value of 0 indicates that gene expression was not detected.

97 4.2.4 ' Correlation of gene array and protein expression data in dendritic cells

To evaluate the significance of the gene array profiles, a number of chemokines and receptors, some known to be modulated on DC maturation, were assessed at the protein level. Transcript data was taken from the microarray data discussed in Section 4.2.2 and further detailed in Appendix 1.

Surface expression of chemokine receptors, on both day 9 immature and mature DCs, was assessed by FACS analysis. For measurement of chemokine levels, day 9 immature and mature DCs were washed thoroughly, re-plated either with or without IFN-y + / - Staphylococcus aureus Cowan’s strain (SAC) or lipopolysaccharide (LPS) and supernatants removed for determination of

chemokine levels by ELISA at 4 and 48 hours. This allowed evaluation of both the constitutive and inducible potential of the immature and mature DC populations to produce chemokines at the protein level.

The array analysis demonstrated that CCRl and CCR5 mRNA expression was decreased in the mature DC population, while CCR7 expression increased in mature DCs. Consistent with this result: the percentage of CCRl- and CCR5-positive cells was shown to decrease in the mature DC population, while that expressing CCR7 was increased, as demonstrated by surface

immunofluorescent staining and flow cytometry (Fig. 4.6).

CCL13, CCL15 and CCL23 all showed decreased transcript expression and decreased protein production in mature DCs (Fig 4.7). Immature dendritic cells constitutively produced all three of

these chemokines, although only at a very low level for CCL15. IFN-y + LPS treatment increased the levels of CCL13 and CCL15 produced, whilst lowering the amount of CCL23 synthesised. IFN- y + SAC increased the amount of CCL15 produced, although not to the level of IFN-y + LPS,

whereas the amounts of CCL13 and CCL23 were decreased.

98 CCL3, SCYA3L and CCL4 mRNA transcripts were decreased on maturation, CCL3 and CCL4

protein levels were also decreased on DC maturation, SCYA3L protein levels were not examined

due to lack of available reagents (Fig. 4.8). In all three cases mature DC transcript levels were very

low. The production of these chemokines by mature DCs, constitutive or induced, was either below

the limit of detection or at very low levels. Interestingly, CCL3 and CCL4, although exhibiting high

transcript levels in immature DCs, both required induction by IFN-y + LPS or IFN-y + SAC, before protein levels could be detected in the supernatants (Fig. 4.8).

CCL17 and CCL22 transcript and protein levels both increased on maturation (Fig. 4.9). The highest levels of protein were expressed constitutively by mature DCs. Further stimulation of the mature DC population either resulted in similar or, in the case of IFN-y + SAC, decreased levels of chemokines produced. Conversely, CCL18 transcript and protein levels are decreased in mature

DCs, whereas a high level of CCL18 is produced constitutively from immature DCs and further stimulation of the immature DCs either resulted in similar or, in the case of lFN-y+ SAC, decreased levels of chemokine produced.

Finally CCL2 mRNA expression levels increased on maturation of DCs (Fig. 4.10), which is not consistent with the protein production data, where larger increases of CCL2 production were detected following activation of immature DCs.

99 CCR1 mRNA Surface expression of expression in DCs CCR1 by DCs

40-1

I 3 0 H Ü) o CL U 2 0 - o

10-

immature mature immature mature

C C R 5 mRNA Surface expression of ex p ressio n in D C s CCR5 by DCs

30-1

CD 'P. - 30 Ï 20-1 X o Cl LU CO 2 0 O Q ^ 1 0 4

immature mature immature mature

C C R 7 mRNA Surface expression of ex p ressio n in D C s CCR7 by DCs

200n lOOn

> II Ü5 o CL LU CD 100 - Ü 50 $ 3 Q

immature mature immature mature

Figure 4.6 Chemokine receptor mRNA and cell surface expression levels

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. For mRNA expression, RNA was extracted and reverse transcribed into ” P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 experiments. For cell surface protein expression, cells were analysed by flow cytometry; data are means + SEM of 3 experiments.

100 CCL13 mRNA expression in DCs Production of CCL13 by DCs

300n 20-1 T E O) l ! 200 - c _l” 10- o 100- o PC C CD E

n- immature mature Control IFN^ + LPS IFN^ + SAC

CCL15 mRNA expression in DCs Production of CCL15 by DCs

100-1 20n

g

10- O U 1

immature mature Control IFN-y + LPS IFN^ + SAC

Production of CCL23 by DCs CCL23 mRNA 200 expression in DCs 50-1 immature DCs C— ] mature DCs If CD LU 2 100 w 25 O C L CD O E ^ 1 £ l immature mature Control IFN^ + LPS IFN^ + SAC

Figure 4.7 CCL13, CCL15 and CCL23 mRNA and protein levels in DC populations

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. For mRNA expression, RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means ( t-SEM) of 6 experiments. For chemokine production levels, day 9 immature and mature DCs were washed thoroughly and incubated for a further 48 hours with the appropriate stimuli. Chemokine levels were measured by ELISA. Data are means (+SEM) of 3 experiments.

101 CCL3 mRNA expression in DCs Production of CCL3 by DCs

75-1 75n

P n 50- 50-

LU CD O O 254

immature mature Control IFN-y + LPS IFN-y + SAC

CCL4 mRNA expression in DCs Production of CCL4 by DCs 50-1

Pn ~ 30

L U CD 2 0 2 5 -

cr CD

immature mature Control IFN-y + LPS IFN-y + SAC

SCYA3L mRNA expression in DCs

300n

P 3 2 0 0 -

UJ CD

immature mature

Figure 4.8 CCL3, CCL4 and SCYA3L mRNA and protein levels in DC populations

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. For mRNA expression, RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 experiments. For chemokine production levels, day 9 immature and mature DCs were washed thoroughly and incubated for a further 48 hours with the appropriate stimuli. Chemokine levels were measured by ELISA; appropriate reagents were not available for SCYA3L. Data are means (+SEM) of 3 experiments.

102 CCL17 mRNA expression in DCs Production of CCL17 by DCs

750-1 300n i f 2 3 500- o) 2 0 0 - o . >,

250- 100-

immature mature Control IFN-y + LPS IFN-y + SAC

CCL22 mRNA expression in DCs Production of CCL22 by DCs

400n 400n

E 3004 O) c ^ 200 H 1 2 200 - o I o

CO 1004 100-

— Control IFN^ + LPS IFN-y + SAC immature mature

CCL18 mRNA expression in DCs Production of CCL18 by DCs

750n 750n C— 3 immature DCs I EH123 mature DCs Ï § 500- g) 500- X

a:p CD 250- 8 250 E ^ n immature mature Control IFN-y' + LPS IFN-y + SAC

Figure 4.9 CCL17, CCL22 and CCL18 mRNA and protein levels in DC populations

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. For mRNA expression, RNA was extracted and reverse transcribed into ^'^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by nonnalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 experiments. For chemokine production levels, day 9 immature and mature DCs were washed thoroughly and incubated for a further 48 hours with the appropriate stimuli. Chemokine levels were measured by ELISA. Data are means (+SEM) of 3 experiments.

103 CCL2 mRNA expression in DCs Production of CCL2 by DCs

40 0 n 7 . 5 - 1 immature DCs mature DCs

— i — % ~ 3 0 0 - 1 ^ 2 3 F ^ 5 . 0 - c X ^ CM lu 2 200 _ l O P ■ ' o 2 . 5 - E 100 X X

1 1 0 . 0 - immature mature Control IFN-y + LPS IFN-y + SAC

Figure 4.10 CCL2 mRNA and protein levels in DC populations

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. For mRNA expression, RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 experiments. For chemokine production levels, day 9 immature and mature DCs were washed thoroughly and incubated for a further 48 hours with the appropriate stimuli. Chemokine levels were measured by ELISA. Data are means (+SEM) of 3 experiments.

4.2.5 Chemokine and chemokine receptor expression in osteoclasts

In most cases, very little variation was seen between the chemokine and chemokine receptor mRNA transcript levels in osteoclasts versus monocytes, suggesting that expression of these genes is not modulated upon differentiation of monocyte precursors to osteoclasts. However, some chemokine and chemokine receptor transcript levels, such as CCL3, SCYA3L, CCL4, CCL20, CXCL 8 and

CCR2 showed down modulation during the differentiation process (Fig. 4.11). These events were not unique to this differentiation process as these genes were also downregulated upon differentiation of monocytes to DCs. Yet one chemokine, CXCL5, was notably upregulated in osteoclast preparations, suggesting an important role for this chemokine in osteoclast biology (Fig

4.12).

104 CCL3 mRNA in SCYA3L mRNA In monocytes and monocytes and osteoclasts osteoclasts

400-1 400

.9 'w 3004 2 3 Q. 9- ^ 2 200 - LU 2 2 0 0 - z f 0 : CO — 100- E ^ 100-

monocyte osteoclast monocyte osteoclast

CCL20 mRNA In monocytes and CCL4 mRNA In osteoclasts monocytes and osteoclasts 50-1 300n

c 3 3 2 0 0 - D. Q. 2 2 5 -

CO 100-

monocyte osteoclast monocyte osteoclast

CXCL8 (IL-8) mRNA In CCR2 mRNA In monocytes and monocytes and osteoclasts osteoclasts

300n 50n

3 2 0 0 - Q. Q. 2 2 5 -

100-

monocyte osteoclast monocyte osteoclast

Figure 4.11 CCL3, SCYA3L, CCL20, CCL4, CXCL8 and CCR2 gene expression in monocytes and osteoclasts Monocytes were purified from PBMC and osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 (monocytes) or 7 (osteoclasts) experiments.

105 CXCL5 mRNA in monocytes, osteoclasts, immature and mature DCs

75 lî 2 3 50

DUP ca 2 5 - E ^

monocyte osteoclast immature DC mature DCs

Figure 4.12 CXCL5 gene expression in monocytes, osteoclasts and DC Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature and immature respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P- labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 (monocytes) or 7 (osteoclasts) experiments.

4.2.6 7TM GPCRs expressed in monocytes, osteoclasts and dendritic cells

In contrast to the chemokine receptors, a number of GPCRs show downregulation on differentiation of monocytes to osteoclasts as demonstrated in Figure 4.13. EDG 6 , EMRI, CD97 and P 2 Y2 transcript levels are all decreased in the osteoclast preparations. This is not unique to the monocyte

—> osteoclast axis, as all four genes are also decreased in immature DCs. Interestingly, P 2 Y 2 increased again on maturation of DC. The dendritic cell preparations showed examples of differential gene expression unique to this arm of myeloid differentiation, with downregulation of

C3AR, C5aR and C3aR on mature DCs (Fig. 4.14). H963, an orphan GPCR, showed increased expression uniquely in mature DCs. GPR105 demonstrated decreased expression in mature DCs

106 when compared to immature DCs, the transcript levels were also very low in osteoclasts, and more significantly, the DC precursors, monocytes (Fig. 4.15). This result suggested that expression of

GPR105 is unique to immature DCs amongst these four cell types.

P2 Y2 mRNA in EDG 6 mRNA in monocytes and monocytes and osteoclasts osteoclasts

30n 50-1

9- >< X LU ro 25-

monocyte osteoclast monocyte osteoclast

EMR1 mRNA in CD97 mRNA in monocytes and monocytes and osteoclasts osteoclasts

30n 150 II 125 II 100 75

0p ^ CD 50 E ^ 25

0 monocyte osteoclast monocyte osteoclast

Figure 4.13 Differential GPCR gene expression in monocytes and osteoclasts Monocytes were purified from PBMC. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 (monocytes) or 7 (osteoclasts) experiments.

107 C3aR mRNA C5aR mRNA expression in DC expression in DC

% -5 30

LU CD 2 0 LU CD 2 5

immature mature immature mature

C3AR mRNA H963 mRNA expression expression in DC in DC 75 n

% 30 II 5 0 -

LU CD 2 0 Q: CD 2 5 - E ^

immature mature immature mature

Figure 4.14 Differential GPCR gene expression in DC populations

Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 DCs were ineubated for 72 hours with or without MCM (mature and immature respectively). RNA was RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 experiments.

108 3 0 0 i

LU CD

immature DC mature DC monocytes osteoclasts

Figure 4.15 GPR105 mRNA expression in myeloid cells Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated for 72 hours, with or without MCM (mature and immature respectively). Osteoelasts were derived from monocytes in the presenee of M-CSF and RANKL for 9 days. RNA was extraeted and reverse transcribed into ^^P- labelled cDNA, which was hybridized to a gene microarray. Expression levels were ealculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of 6 (monocytes, immature DCs and mature DCs) or 7 (osteoclasts) experiments.

109 4.3 Discussion

The use of microarray technology to elucidate gene expression profiles is becoming commonplace.

The aim of this chapter was not to generate complete profiles for the cell types investigated,

requiring labour intensive, quantitative real time PCR for confirmation, but to look at differential

expression patterns in specific gene families, to identify genes of interest.

Day 9 immature and mature (MCM treated) DC populations, prepared for RNA extraction and

microarray profiling, were characterised for surface marker expression by flow cytometry.

Immature DCs were HLA-DR (MHC class II)-medium, CDSO-low, CD 8 6 -negative and CD83-

negative; mature DCs were HLA-DR-high, CD80-high, CD 8 6 -high and CD83-positive. This is

characteristic of immature and mature DCs (Banchereau and Steiijman, 1998; Reddy et ah, 1997;

Spisek et ah, 2001). More extensive analysis of these DC populations is described Chapter 3.

Monocyte populations prepared for microarray profiling yielded high purity of CD 14 expressing

cells. Day 9 osteoclast preparations yielded populations where approximately of 65% of the nuclei present could be attributed to multinucleated cells, of which over 90% were TRAP-positive.

Although the populations were clearly not homogenous, a high degree of the nuclear material could be attributed to the osteoclast like cells. Boissy et ah have provided evidence that the nuclei of resorbing osteoclasts isolated from embryonic chick long bones, or differentiated in vitro from murine spleen cells, are all transcriptionally active (Boissy et ah, 2002). This evidence suggests assessing the percentage of nuclei attributed to multinucleated cells will correlate with the contribution of these cells to transcriptional activity. RNA was extracted from these osteoclast preparations and all demonstrated message for RANK and calcitonin receptor by RT-PCR.

Expression of both these receptors is necessary to allow description of these as osteoclast-like cultures. It should be noted that calcitonin receptor mRNA cannot be detected in macrophage polykaryons (Arai et ah, 1999; Takahashi et ah, 1995). Neither CTR nor RANK mRNA transcripts

110 were detected in the precursor monocyte population. As discussed in Chapter 3, RANK has previously been detected on CD 14-positive peripheral blood monocytes (Shalhoub et ah, 2000), but the inability to demonstrate RANK transcripts in the three monocyte samples tested suggests an increase of RANK message in the osteoclast samples. These characteristics are consistent with in vzfm-derived osteoclast cultures previously described (Quinn et ah, 1998; Flanagan and Massey,

2003). More extensive characterisation of these osteoclast populations can be seen in Chapter 3.

Formaldehyde gel electrophoresis and ethidium bromide staining was used to evaluate the integrity and size distribution of total RNA. Each sample demonstrated 18S and 28S ribosomal RNA as sharp bands on a stained gel. These RNA samples were then used to make labelled cDNA probes with which to probe the custom made microarray. Each cDNA sample was hybridised to a microarray and transcript expression data obtained.

The transcript expression data acquired from the microarray profiling, was normalised to a panel of forty housekeeping genes and subjected to hierarchical clustering using the Genespring software.

Data are grouped together solely on the similarity between gene expression profiles. This analysis resulted in the formation of four clearly defined clusters, correlating with the cell populations of immature DCs, mature DCs, monocytes and osteoclasts. This relationship between cell specificity and similarity of expression profiles suggests that patterns of transcript changes between cell types are present.

For most genes, changes in mRNA abundance relate to changes in protein abundance. However, during the microarray experiments mRNA is reverse transcribed to cDNA, this is necessary as RNA is prone to being destroyed and therefore difficult to work with. A problem with cDNA production is that not all mRNAs are reverse transcribed with the same efficiency. This leads to reverse transcription bias, which can change relative amounts of different cDNA measured in the same sample. However, when comparing the same gene across two cell populations this should not be a

111 problem, unless the mRNA is not transcribed at all. Another problem caused by bias is that mRNA may be transcribed for only part of their lengths, making them less likely to bind and stay bound to

their compliments on the array. Again this phenomenon is likely to remain consistent across cell populations when looking at the same gene. With this in mind it should be noted that all expression

levels are arbitrary and will depend in part on the transcription efficiency from the mRNA and in part on the design of the 500 - 900 base pair UTR sequences that comprise the microarray.

Therefore the most reliable data obtained from a microarray is comparative data, where obvious

changes in the level of expression of a particular gene are seen.

Differentially expressed genes were highlighted where an expression level of greater or less than 3

times that seen in the other cell types could be attributed to one cell type. Most of the genes selected

by this analysis are chemokines and chemokine receptors where differential expression can be

attributed to one of the dendritic cell populations, many of which will be considered on validation

of the microarray. All genes selected in this process will be discussed in the relevant section.

The use of a custom made array allowed concentration on specific areas of interest for the cell types

investigated, but also required an element of validation in order to gain confidence in the

information obtained. Examining the chemokine and chemokine receptor data obtained for dendritic

cells and comparing with a study of protein production and surface expression in these cells, and the

body of literature available in this field, allowed this validation.

Microarray gene expression profiles of immature and mature DCs revealed patterns of chemokine

and chemokine receptor expression that distinguish both stages of maturation. When compared to

mature DCs, immature DCs exhibit higher levels of expression of CCL3 (MlP-la), CCL4 (MIP-

IP), CCL13 (MCP-4), CCL15 (MlP-5, HCC-2), CCL18 (PARC, MlP-4), CCL23 (Ckp- 8 , MPlF-1),

CCRl and CCR5, while mature DCs exhibit increased expression of CCL2 (MCP-1), CCL17

(TARC), CCL22 (MDC) and CCR7.

112 Expression of particular chemokine receptors has previously been associated with DC development,

including CCR7 with mature DCs and CCRl, CCR5 and CCR 6 with immature DCs (Yang et ah,

1999; Power et ah, 1997; Greaves et ah, 1997; Sallusto et ah, 1998; Sozzani et ah, 1997). As

expected, CCR7 markedly increased and CCRl and CCR5 decreased significantly on maturation.

Cell surface levels of these chemokine receptors, as analysed by immunofluorescent staining and

flow cytometry, correlated well with the transcript levels. Significant levels of CCR 6 were not

observed in these in vzYro-derived DCs (data not shown), consistent with several reports that

monocyte-derived DC do not express CCR 6 , unlike CD34+ progenitor-derived DCs (Yang et ah,

1999; Power et ah, 1997; Greaves et ah, 1997). Expression of CCRl and CCR5 by immature DCs is

appropriate in their role as immune sentinels, residing within peripheral tissues and sampling the

microenvironment for foreign antigens. CCRl and CCR5 expression are consistent with their requirements to migrate to areas of inflammation, where ligands for these receptors, such as CCL3,

CCL4, CCL5 (RANTES) and CCL15 are expressed (Pashenkov et ah, 2002; Sozzani et ah, 1997).

Coordinated changes in chemokine receptor expression on DC maturation, including downregulated

expression of CCRl and CCR5 and upregulated expression of CCR7, allows the chemotactic recruitment of DCs to T cell rich areas of the lymph nodes that express the CCR7 ligands CCL19

and CCL21 (Baekkevold et ah, 2001). Here they can interact with naive T cells and initiate an

immune response.

The study in this chapter looks at constitutive and inducible chemokine production by immature and mature DCs. The stimuli used to induce chemokine production fi"om DCs were IFN-y plus either

LPS or SAC. IFN-y is a potent Thl cytokine, associated with strong cellular immune responses and inflammation, which is produced by T cells in response to contact with monocyte-derived DCs and their products (Romagnani, 1999; Spellberg and Edwards, Jr., 2001). Thus, monocyte-derived DCs are generally considered to be Thl polarising (Zou et ah, 2000) and as such, would experience high levels of IFN-y on contact with naïve T cells. With this in mind, IFN-y was selected as a stimulus

113 for chemokine production by in v/Yro-derived immature and mature DCs, along with classic Thl inducing agents, LPS and SAC (Cameron et ah, 2001).

Changes in production of chemokines on DC maturation have been widely reported (Wan and

Bramson, 2001; Rescigno et ah, 1999). Expression of CCL13 by human in vitro monocyte-derived

DCs has been previously noted (Hashimoto et ah, 1999), but the effect of maturation on regulation has not been studied extensively. Transcript and protein data in this chapter confirmed the expression of this molecule in DCs and demonstrated a downregulation on maturation. CCL13 is a chemoattractant for monocytes, activated T cells and NK cells, working through the CCR2 receptor.

As such it is widely expressed at disease sites and fimdamental in the induction of cellular infiltrates

(Miotto et ah, 2001; Taha et ah, 2000). This role is concurrent with the presence of immature DCs at these sites. CCL23 is reportedly constitutively produced by monocyte-derived DCs and its expression downregulated by inflammatory mediators (Nardelli et ah, 1999a), as was demonstrated in the transcript and protein analysis. Expression of CCL15 by DCs has not been previously reported but, intriguingly, this chemokine has high protein homology and functional similarities with CCL23 (Youn et ah, 1997b), CCL15 also showed reduced transcript and protein expression in mature DC. Both CCL15 and to a greater degree CCL23 protein are produced constitutively from immature, but not mature, monocyte-derived DC. Interestingly, stimulation of immature DCs with

IFN-y and either LPS or SAC resulted in much increased levels of CCL15, but reduced levels of

CCL23. Both CCL15 and CCL23 have been reported to bind to CCRl, and CCL15 also acts through CCR3 (Nardelli et ah, 1999b; Forssmann et ah, 2001). Both chemokines are implicated in inducing migration of neutrophils, monocytes and lymphocytes to sites of inflammation via these receptors (Wang et ah, 1998; Forssmann et ah, 1997). However, the differing expression patterns of these proteins observed in DC subpopulations may suggest a new functional distinction between these ligands in the sequence of an immune response.

114 Other inflammatory chemokines such as CCL3 and CCL4 are reported to be produced at high levels in DCs in response to maturation signals, however the increase is transient, lasting only for a few hours (Sallusto et ah, 1999; Fischer et ah, 2001). Transcript levels for these chemokines were markedly reduced in mature DCs, probably a reflection of analysing the cells 72 hours after addition

of the maturation stimulus. Indeed, upon stimulation with IFN-y and LPS and in particular, IFN-y and SAC, both CCL3 and CCL4 protein levels were upregulated. Whereas most other chemokines

associated with immature DCs showed constitutive protein production by these cells, CCL3 and

CCL4 required a stimulus for any protein levels to be detectable. This is concurrent with these

chemokines being widely described as inducible chemokines, with a very well defined role in the build up of cellular infiltrates at disease sites; working through the receptors CCRl, 3 and 5 and

calling in macrophages, lymphocytes, eosinophils, dendritic cells and NK cells (Driscoll, 1994;

Cook, 1996). Production of these chemokines only on activation of immature DCs would ensure

that inflammatory infiltrates only occur where required and only in response to signals, such as

IFN-y, that indicate true immune T cell activation. Furthermore, fully mature DCs producing these

chemokines would not be appropriate as this would result in inflammatory infiltrates at the thymus

and lymph nodes. There is a second non-allelic gene for CCL3, gene designation SCYA3L (Nakao

et ah, 1990). The product of this gene has been identified and functionally characterised, it shares

90% homology with CCL3 and enhanced receptor interactions (Nibbs et ah, 1999). Unsurprisingly,

the transcript for this gene showed a very similar pattern of expression to CCL3 in inunature and

mature DCs. Protein levels of SCYA3L were not evaluated due to lack of available reagents.

In this study, CCL17 and CCL22 both showed increased transcript levels in mature DCs and high

levels of constitutive protein production from these cells. This is consistent with reports that CCL17

levels are upregulated on maturation (Sallusto et ah, 1999; Vissers et ah, 2001). The high level of

CCL22 transcript observed upon maturation is consistent with DCs as a major source of this

chemokine, both in vitro and in vivo (Vulcano et ah, 2001). CCL17 and CCL22 are both

115 constitutively expressed in the thymus and share utilisation of the chemokine receptor CCR4.

CCL17 is thought to guide the migration of positively selected thymocytes. In contrast, CCL22, while capable of attracting thymocytes, also plays a role in attracting monocyte-derived DCs and activated NK cells, suggesting that it may act on an additional receptor (Bochner et ah, 1999;

Schaniel et ah, 1999). Both CCL22 and CCL17 have been implicated in the attraction of activated T cells and the formation of lymphocyte-dendritic clusters in inflamed skin and secondary lymphoid tissues (Katou et ah, 2001). The production of CCL19 by mature DCs may also have been expected, however the transcript levels of this chemokine in mature DCs were unremarkable. This may be a reflection of the time-point evaluated, as CCL19 production has been demonstrated in mature DCs only at late time-points (Sallusto et ah, 1999).

CCL18 is also expressed at high levels in the lymph nodes, and preferentially attracts naive T cells,

B lymphocytes and immature DCs (Adema et ah, 1997; Hieshima et ah, 1997). No receptor for

CCL18 has been identified to date. In concordance with the transcript and protein data seen here,

CCL18 production from immature DCs represents one of the most abundantly secreted chemokines by DCs and is downregulated on maturation (Vulcano et ah, 2003). In this study, immature DCs secreted very high constitutive levels of CCL18, with little change on activation with IFN-y and

LPS or IFN-y and SAC, but vastly reduced levels from mature DCs under all conditions. There is evidence that stimulation of naïve T cells with immature DCs induces the generation of lL-10 producing T regulatory cells (Jonuleit et ah, 2000). CCL18 is chemotactic for both immature DCs and naïve T cells and, as such, a regulatory role for CCL18 production by immature T cells has been postulated (Vulcano et ah, 2003).

Consistent with the transcript data in this chapter, the proinflammatory chemokine CCL2 is reported to be produced upon maturation of DCs in a sustained manner (Sallusto et ah, 1999). However, production of CCL2 protein from LPS and IFN-y stimulated immature DCs was significantly higher

116 than from mature DC. As discussed earlier, with respect to CCL3 and CCL4, this may be a reflection of the timepoint chosen for study. The culture supernatants were taken 48 hours after treatment with the stimuli, if CCL2 is an early indicator of DC maturation, mRNA may be transcribed, translated and indeed degraded in this timescale.

During the analysis of chemokine protein production from DCs, four hour timepoints were also evaluated. CCL3, CCL4 and CCL15 levels could not be detected at four hours (data not shown).

CCL2, CCL13, CCL17, CCL18, CCL22 and CCL23 demonstrated lower levels of protein, with the same patterns as the 48-hour timepoint (data not shown).

The purpose of this study was to demonstrate a correlation between transcript and protein levels in

DC populations. With the exception of CCL2, protein levels could be seen where mRNA was present, if not constitutively, then on activation. These data suggest that expression profiles demonstrated by the array profiling are representative of protein production by immature and mature DC. The study also describes for the first time, expression of CCL15 by monocyte-derived

DC. Whilst evaluating chemokine profiles with relation to dendritic cell biology would require more extensive assessment, the gross picture fits in well with DC biology and the plethora of literature on this topic.

Major changes in DC chemokine and chemokine receptor expression profiles are seen, and can, for the most part, be related to the function of these cells. What then of changes in these entities on the differentiation of osteoclasts from their monocyte precursors? The differences between the expression of chemokine and chemokine receptor transcripts in osteoclasts and monocytes are mostly unremarkable, perhaps a reflection of the relative unimportance of these molecules in the biology of osteoclasts, for which the main function is bone resorption, which is carried out in situ at the site of differentiation from their precursors. Transcript levels for CCL3, the non-allelic CCL3

gene SCYA3L, CCL4, CCL20, CXCL 8 (lL-8 ) and CCR2 were decreased in osteoclasts. These

117 events are not unique to this differentiation pathway, as all were markedly decreased in immature

and mature DCs as well, suggesting that the significance of these profiles is the important role of

these chemokines and receptors in monocyte biology.

CCL3 and CCL4 are widely reported as produced from monocytes and macrophages (Wolpe et ah,

1988) concurrent with the high levels of transcript seen in the monocyte samples (Bug et ah, 1998).

CCL3 and CCL4 are still expressed at levels in the osteoclast preparations that may be significant,

as transcript levels were similar to those seen in immature DCs, and high levels of CCL3 and CCL4

protein were demonstrated on activation of these cells. CCL20 (MlP-3a) is another inflammatory

chemokine, the major role of which is attracting Langerhans cells, B cells and neutrophils to the site

of infection (Caux et ah, 2000; Akahoshi et ah, 2003; Liao et ah, 2002), as such it is hard to postulate a role for this chemokine in osteoclast biology. Maybe the remaining low levels of mRNA transcript are due to the contaminating mononuclear presence in the osteoclast cultures. CCR2 is

expressed on monocytes and has been shown to regulate their recruitment to sites of inflammation

(Han et ah, 1998). Therefore the decreased expression of the CCR2 transcript in osteoclast and dendritic cells could be a loss of requirement for expression of this receptor in these cell types.

Indeed CCR2 has been shown to be downmodulated in murine bone marrow cultures on treatment with RANKL (Lean et ah, 2002).

Two further genes differentially upregulated in the monocyte population were CX 3 CRI and hm74.

CX3 CRI, the fractalkine receptor, is expressed on CD 16-positive monocytes (activated monocytes expressing the Fc receptor CD 16), implicating fractalkine as a mediator of monocyte chemotaxis into tissues (Ancuta et ah, 2003). CX3CR1 is reportedly expressed on murine osteoclast cultures

(Lean et ah, 2002). hm74, a seven transmembrane GPCR for nicotinic acid (Tunaru et ah, 2003) was originally cloned from a human monocyte cDNA library (Nomura et ah, 1993), indicating the likely expression of the genes in monocytes. These inconclusive results highlight the need for

118 confirmation of transcript expression by RT-PCR, Northern blots or indeed, measurement of protein expression.

Lean et al. describe the only other extensive study of chemokine ligand and receptor expression in osteoclasts by quantitative RT-PCR (Lean et ah, 2002). In their study, CCL9 (MIP-ly) and its receptor CCRl are postulated to be the major chemokine ligand / receptor species expressed by murine osteoclasts. The study detailed in this chapter is the first, to my knowledge, to examine human osteoclast cultures for chemokine and chemokine receptor gene expression. While there is no human homologue of the CCL9 gene, this study demonstrated expression of CCRl transcript in human osteoclast cultures, with levels comparable to that seen in the immature DC population. In addition, CCL3 and CCL4, demonstrated transcript levels similar to that of immature DC. In the

DCs, these levels of transcript resulted in measurable protein production. Unfortunately, protein production was not examined in osteoclasts due to limited cell numbers obtained. However, it is interesting to speculate that these alternative CCRl ligands replace the requirement for CCL9 in human osteoclasts.

CXCL8 (IL-8 ), a CXC chemokine with a major role to play in inflammation, is produced by many cell types, including monocytes (Harada et al, 1994; Mukaida et al, 1998). Despite the

downregulation of mRNA seen in dendritic cells and osteoclasts, CXCL 8 production from these cell types has also been sited, at least in diseased tissues (Zhu et al, 2000; Rothe et al, 1998). Whereas

CXCL5 (ENA-78) the only gene from our array showing differential upregulation in the osteoclast population, has not been previously demonstrated in osteoclasts. Expression of CXCL5 in osteoclasts is evaluated further in Chapter 6 .

The panel of other GPCR ligands included on the array showed unremarkable expression patterns, with no differential expression seen between cell types. The G protein-coupled receptors and orphan

GPCRs; EDG 6 , P2 Y2 , CD97 and EMRI, showed some downregulated expression on

119 differentiation of monocytes to osteoclasts. As previously, this downmodulation was not unique to this arm of the differentiation pathway, as expression also decreased on differentiation to DCs. Both

EMRI and CD97 message have been reported in cells of the myelomonocytic lineage. Indeed,

EMRI is predicted to be the human ortholog of the macrophage restricted F4/80 molecule

(McKnight and Gordon, 1998). These molecules are members of the epidermal growth factor (EGF)

7TM family, expressing large extracellular regions containing a variable number of EGF domains.

EDG 6 , a member of the endothelial differentiation gene (EDG) family of proteins, has been associated with high levels of expression in lymphoid tissue, including peripheral leukocytes, and spleen (Graler et ah, 1999). Indeed, expression of EDG 6 on monocytes and their subsequent activation in response to the ligand for EDG 6 , sphingosine-1 -phosphate, has been reported (Fueller et ah, 2003). Monocytes have been demonstrated to express P 2 Y2 receptors (Jin et ah, 1998); therefore the higher levels of transcript seen in this cell type are not surprising. Interestingly, the transcript analysis showed increased levels of P 2 Y2 on maturation of DCs. P 2 Y2 has been demonstrated on DCs (Berchtold et ah, 1999), but the effect of maturation needs further investigation. This study therefore showed very little difference in the transcript analyses of GPCRs between monocytes and osteoclasts. However, the GPCRs and orphan GPCRs, showed some differential expression patterns worthy of note in DCs. The C3 and C5 anaphylotoxin receptors

(C3aR, C5aR) showed downregulation of expression on mature DCs, with similar expression levels on monocytes, osteoclasts and immature DC. Dendritic cells generated from human blood monocytes and CD la-positive dendritic cells derived from skin have been shown to express these anaphylotoxin receptors (Kirchhoff et ah, 2001). Furthermore, the same study showed downregulation of the anaphylotoxin receptors and a loss of activity to their respective ligands on maturation, in agreement with the transcript expression data detailed in this chapter. The losses of activity to the anaphylotoxins are concurrent with their role in inflammation and the requirement of immature, but not mature DCs at these sites. Not surprisingly, C3AR (AZ3B) showed the same expression pattern on these cell types as C3aR and C5aR. The C3AR receptor cDNA was isolated

120 with a probe derived from the N-formyl peptide receptor, belonging to the same family as the C3aR and C5aR and shows high homology to this family of receptors (Roglic et al, 1996). H963, an orphan GPCR, showed upregulation of message on the mature DC population. This GPCR has no identified function to date (Jacobs et al., 1997), but this finding of expression in mature DCs might suggest a role for H963 in antigen presentation or T cell activation.

The most exciting observation was the strong increase of the GPR105 transcript in immature DCs.

Immature DCs express GPR105 at substantially higher levels than mature DCs, with expression by monocytes and osteoclasts similar to that of mature DCs. This G protein-coupled receptor GPR105

(KIAAOOOl), a relative of the P2Y family of receptors (Chambers et al, 2000), recently renamed as the P 2 Yi4 receptor (Abbracchio et al, 2003), was first cloned from an immature myeloid cell line

(Nomura et al, 1994). No function has yet been attributed to this receptor, although the rat ortholog,

VTR 15-20, is regulated by in vivo challenge with LPS and has been implicated in neuroimmune function (Charlton et al, 1997). Additional expression studies and possible functional roles of

GPR105 by DCs are investigated in Chapter 5.

This chapter sought to elucidate differential transcript expression in monocytes, osteoclasts, immature and mature DCs of chemokines, chemokine receptors, other seven transmembrane

GPCRs and their known ligands. Firstly the array was validated by an extensive study comparing chemokine and chemokine receptor transcript and protein levels in immature and mature DCs, this data has also been discussed here in the context of the plethora of relevant literature available. In particular, the presence of CCL13 in immature, but not mature DCs, has been confirmed at the protein level, and the expression of CCL15 by DCs (previously unreported), is also confirmed at the protein level. Other differentially expressed genes have been highlighted and discussed. Perhaps the most interesting of these, the high levels of GPR105 transcript in immature DCs and the increase in

CXCL5 transcript on differentiation of monocytes to osteoclasts, will be investigated in detail in the following chapters.

121 Chapter 5

GPRlOSinDCs

122 Chapter 5

5.1 Introduction

The dendritic cell array data discussed in Chapter 4, revealed an interesting expression profile of the

G protein-coupled receptor, GPR105 (previously termed KIAAOOOl). The cDNA encoding

GPR105, a relative of the P2Y family of receptors (Chambers et ah, 2000), and more recently renamed as the P 2 Yi4 receptor (Abbracchio et ah, 2003), was first cloned from an immature myeloid cell line (Nomura et ah, 1994). Uridine 5’-diphosphoglucose (UDP-glucose) was later

identified as a potent agonist ligand of the GPR105 molecule (Chambers et ah, 2000). No

physiologic function has yet been attributed to this receptor, although the rat ortholog, VTR 15-20,

is regulated by in vivo challenge with LPS and has been implicated in neuroimmune function

(Charlton et ah, 1997). The rat and mouse orthologs for GPR105, exhibiting 81 and 83% amino acid

identity respectively to the human protein, both respond to the putative ligand UDP-glucose

(Freeman et ah, 2001), however the existence of other physiological ligands remains a possibility.

Since the demonstration of non-lytic release of ATP from stressed (mechanically agitated) cells

(Bumstock, 1997)’ nucleotides have emerged as a family of extracellular signalling molecules with

wide ranging effects. For example, ADP induces platelet aggregation, through binding the P2Yn

receptor (Hollopeter et ah, 2001), and ATP mediates chemotaxis, cytotoxic responses and cytokine

secretion through the P2X? receptor (North, 2002). As dendritic cells express many P2Y and P2X

receptors (Berchtold et ah, 1999), there is a possibility that purine or pyrimidine-derived molecules

influence dendritic cell maturation. Indeed extracellular ATP has demonstrated co-stimulatory

effects on DCs, synergising with TNF-a to induce DC maturation (Berchtold et ah, 1999; Schnurr

et ah, 2000) via the P2Yn receptor (Wilkin et ah, 2001).

123 Considering the regulation of the rat ortholog of GPR105, VTR 15-20, on in vivo challenge with

LPS, the emerging role of extracellular nucleotides in modulating the action of LPS is of great interest. For instance, extracellular ATP has been shown to enhance TNF-a message levels in the

LPS treated murine macrophage cell line RAW 264.7 (Tonetti et ah, 1995) and increase IL-ip release in LPS-treated human macrophages (Ferrari et ah, 1997). An ATP analogue has even been shown to protect mice from a lethal LPS challenge (Proctor et ah, 1994). Indeed stimulation of

LPS-matured dendritic cells with ATP triggered the release of IL-lp and TNF-a, probably via the

P2X? receptor (Ferrari et ah, 2000). Furthermore, LPS binding regions have been demonstrated on the nucleotide receptor P2X?, suggesting that nucleotide receptors may directly modulate the action of LPS (Denlinger et ah, 2001). Little is known of the signalling capabilities of nucleotide sugars such as UDP-glucose. However, regulated release of cellular UDP-glucose has been demonstrated, implicating it as an extracellular signalling molecule (Lazarowski et ah, 2003). Induction of DC maturation through exposure to necrotic tumour cells (Sauter et ah, 2000) has lead to the proposal that “danger signals” (Gallucci and Matzinger, 2001) could have a major role in the initiation of an immune response. It is an attractive hypothesis that these and other small molecules may be released as “danger signals” from cells at inflammatory and / or tumour sites, either through passive diffusion following traumatic lysis or through regulated exocytosis.

This chapter aims to confirm the high levels of expression of GPR105 by human monocyte-derived immature DCs, concomitant with the lower levels expressed by mature DCs and monocytes.

Agonist capabilities of UDP-glucose on this receptor using a GPRl 05-expressing HER 293 cell line will be substantiated. The functional effects of UDP-glucose on DCs were also investigated, through measurement of intracellular calcium release and assays for DC maturation and function.

Potential synergy of UDP-glucose with LPS will also be described. Thereafter, possible roles for

GPRl 05 on immature DCs are suggested and discussed further.

124 5.2 Results

5.2.1 GPR105 transcript expression in DCs

The dendritic cell gene array data detailed in Chapter 4 demonstrates elevated GPR105 expression in immature DCs. Expression of this gene is notably absent in monocytes and osteoclasts and very low in mature DCs (Appendix 1, Fig. 4.16). Figure 5.1 confirms that high levels of GPR105 mRNA are present not only in day 9 immature DCs but also in day 6 immature DCs, with much lower levels detected in day 9 mature DCs. Northern blot hybridisation confirms the expression pattern of

GPR105 (Fig. 5.2).

300- day 6 immature DCs % ^ day 9 Immature DCs 2 i Œ >. 2 0 0 day 9 mature DCs uj m < ^ i l 100 E “

Figure 5.1 GPR105 transcript expression in DC populations DCs were derived from monocytes in the presence of GM-CSF and IF-4. RNA was extracted from day 6 DCs or DCs incubated for 72 hours with (day 9 mature) or without (day 9 immature) MCM and reverse transcribed into ^^P-labelled cDNA, which was hybridised to a gene microarray. Resulting transcript expression data was normalised to the average expression level of the housekeeping gene panel. Data are means (+SFM) of 3 separate experiments.

125 donor 1 donor 2 donor 3 A A A ( r r a a. q Q. a. q a a. q Q) f it Q) fi) Q) < < < *< *< SI O) CO c o> CO c o> CO c (D

GPR105

ACTIN

Figure 5.2 Northern blot analysis of GPR105 and P-actin mRNA expression DCs were derived from monocytes in the presence of GM-CSF and IL-4. RNA was extracted from day 6 DCs or DCs incubated for 72 hours with (day 9 mature) or without (day 9 immature) MCM. 5 pg of total cellular RNA was electrophoresed through a 1.2% agarose/16.7% formaldehyde gel, transferred to Zeta-Probe GT membrane and probed with a ^^P-labelled 905 bp GPR105 probe or a ^^P-labelled human actin probe. Hybridisation was performed overnight at 65°C. 3 donors; day 6 immature DCs (lanes 1, 4, 7); day 9 immature DCs (lanes 2, 5, 8 ) and day 9 mature DCs (lanes 3, 6 , 9) are represented.

126 5.2.2 UDP-glucose-induced calcium signalling in a GPRlOS-transfected cell line

An in-house HEK293 cell line co-transfected with the g protein subunit Goti^ and HA-tagged

GPR105 cDNA constructs was used for intracellular calcium flux assays to confirm the agonist effects of UDP-glucose on this receptor. Firstly expression of GPR105 in the cell line was confirmed by immunofluorescent staining and flow cytometry using an antibody to the HA tag (Fig.

5.3). Endothelin was utilised as a positive control for the calcium flux experiments, generating release of intracellular calcium in the non-transfected HEK293 parent cell line. The effects of UDP and glucose were also investigated. Endothelin and UDP showed similar activity on both the

GPRl 05-expressing and parent cell lines, whereas UDP-glucose induced release of intracellular calcium only in the GPRl 05-expressing cell line; glucose had no effect in either cell type (Table

5.1, Fig. 5.4). These data demonstrate specific agonist activity of the GPRl05 receptor in HEK293 transfectants by UDP-glucose.

127 Rat IgG l Rat anti-HA

HEK293 Parent cell line

'I, ...... """fj

10^ 10^ 10'^ 10^ 10" 10 10" 10 log fluorescence log fluorescence

Gai6/GPR105 co-transfectants

'"I ...... 'I, I I im i| I I n n i| I I iin i| I f 1 0 ' 10' 10' 10" 1 0 ' 10" 10" 10 log fluorescence log fluorescence

Figure 5.3 Expression of HA-tagged GPR105 in HEK293 transfectants HEK293 cell lines were analysed by immunofluoresence and flow cytometry for HA-tagged GPRl05 expression. Isotype control antibody profiles are shown in the left panel and anti-HA profiles are shown in the right panel for (A) HEK293 parent and (B) GPRl05/ Gai6 co-transfected cell lines. One representative experiment is shown.

128 HEK293 parent cell line

3 0 0 0 0 - 1

c 2 0 0 0 0 -

X (0 1 0 0 0 0 -

7 5 0 0 T _)

5 0 0 0 - l£

2 5 0 0 -

0 - 1.0x10°®1.0x10-°3 1 1000 1000000

[Agonist] pM

B. G P R l05 / O ai6 HEK293 cell line

3 0 0 0 0 - 1

c Endothelin 2 0 0 0 0 - UDP-glucose X CO 1 0 0 0 0 - UDP G lucose 7 5 0 0 " Z)

LL 5 0 0 0 -

2 5 0 0 -

0 - —I------1 1.0x10-°®1.0x10°3 1 1000 1000000

[Agonist] pM

Figure 5.4 Intracellular calcium release in a GPR105/Gai6 expressing and control HEK293 cell lines HEK293 cells were co-transfected with Gttie and GPRl 05 and a stable cell line was obtained by serial dilution cloning. Calcium responses induced with endothelin, UDP-glucose, UDP and glucose were measured in HEK293 parent (A) and GPRl05 / Gais expressing HEK 293 (B) cell lines using the calcium sensitive dye Fluo4-AM, in the FLIPR. Peak increases in fluorescence units minus basal (F.l.U. Max - Min) are shown; data are triplicates +/- S.D. One representative experiment of 3 is shown.

129 HEK293 parent cell line GPR105/Gai6 HEK293 endothelin 0.94nM 0.7nM

UDP-glucose >3.33|iM 0.46|iM

UDP 276pM 146|xM glucose > ImM > ImM

Table 5.1 EC50 values for various ligands on GPR105/Gai6 expressing and control HEK293 cell lines HEK293 cells were co-transfected with Gaie and GPRl 05 and a stable cell line was obtained by serial dilution cloning. Calcium responses induced with endothelin, UDP-glucose, UDP and glucose were measured HEK293 parent and GPRl 05 / Gai 6 expressing HEK 293 cell lines using the calcium sensitive dye Fluo4-AM, in the FLIPR. Dose response curves are shown in Fig. 5.4 EC50 calculations were performed using Xlfit software.

130 5.2.3 Correlation of UDP-glucose-induced intracellular calcium release with GPR105

expression

As immature DCs, in comparison to mature DCs and monocytes, demonstrated increased expression of GPRl 05 it was vitally important to determine if this particular subset of DCs would respond to the GPRl 05 agonist ligand, UDP-glucose. Calcium signalling in monocytes, immature and mature DCs was measured in response to UDP-glucose (Fig. 5.5). CCL3 served as a positive control for both monocytes and immature DCs expressing CCRl and CCR5, while CCL19 was included as a positive control for mature DCs expressing CCR7. As predicted by the tightly regulated expression of CCRl, CCR5 and CCR7 in these cell types (Sozzani et al., 1998), monocytes and immature DCs both exhibited increased intracellular calcium concentrations in response to CCL3 (EC50 values = 0.025nM +/- 0.0006 SEM (n = 3) for monocytes and 0.22nM +/-

0.07 SEM (n = 4) for immature DCs) while mature DCs showed elevated intracellular Ca^^ concentrations in response to CCL19 (EC50 value = 9 nM +/- 3.9 SEM (n = 4)). No significant response was observed from immature DCs to CCL19, mature DCs showed markedly lower, if not insignificant, responses to CCL3, compared to monocytes and immature DCs. Consistent with the observed expression pattern of GPRl05, immature DCs increased intracellular Ca^^ concentrations in response to UDP-glucose, the GPRl05 agonist ligand (EC50 value = 11.5jLiM +/- 2.1 SEM (n =

4)), while no significant response to this molecule was observed from mature DCs or monocytes.

131 A. Monocytes B. Immature DC

A A 1 0 0 0 0 - 1 10000- 7 5 0 0 - A i

: 5 0 0 0 - . 5 0 0 0 : . I

. 4 0 0 0 - r X 4 0 0 0 - r ( CD L 3 0 0 0 3 0 0 0 -

B 20004 = j 2 0 0 0 - L i.' i f 1000 - 1 0 0 0 - 4- 0 n I I I I I I I I 0 - 10-M0-M0-U0-210-^ 10° 10^ 102 103 1Q4 10-310-^0-310-210-^ 10° 10^ 1Q2 103 10' [Agonist] pM [Agonist] p.M

Mature DC

2 5 0 0 0 - 1

1 5 0 0 0 - c A CCL3

2 5 0 0 0 -

X 4 0 0 0 - r + CCL19 03

3 0 0 0 - UDP-glucose

2 0 0 0 - LL

1 0 0 0 -

0 - - (Jf* 10-310^0-310-210-^ 10° 10^ 1Q2 103 1Q4

[Agonist] pM

Figure 5.5 Intracellular calcium release in response to UDP-glucose in monocytes and DCs Monocytes were purified from PBMC. DCs were derived from monocytes in the presence of GM- CSF and lL-4. Day 6 DCs were incubated for 72 hours with (mature) or without (immature) MCM. CCL3, CCL19 and UDP-glucose-induced calcium responses were measured using the calcium sensitive dye, Fluo4-AM, in conjunction with a FLIPR, in monocytes (A), immature DCs (B) and mature DCs (C). Peak increases in fluorescence units minus basal (F.l.U. Max - Min) are shown. Data are triplicates +/- S.D. One representative experiment of 3 (monocytes) or 4 (immature and mature DCs) is shown.

132 5.2.4 Phenotypic changes in immature DCs in response to UDP-glucose

As GPRl 05 is expressed on immature but not mature DCs and treatment of immature DCs with

UDP-glucose elicited an increase in free intracellular calcium levels, the ability of UDP-glucose to

induce a maturation response from these cells was evaluated. For these experiments TNF-a was

used as a well defined positive control for maturation (Sallusto and Lanzavecchia, 1994) and UDP,

which has no GPRl 05 agonist activity (Chambers et al., 2000), was used as a negative control.

Following treatment, expression of CD80, CD 8 6 , HLA-DR and CD83, which are often cited as

classical markers of DC maturation (Banchereau and Steinman, 1998; Banchereau et al., 2000),

were then quantified by immunofluorescent staining and flow cytometery. Overall, both UDP-

glucose- and TNF-a-treated DCs showed an increase in cell surface HLA-DR, CD 80, and most

strikingly CD 8 6 and CD83 levels, consistent with their acquisition of a mature DC phenotype; while UDP-treated DCs maintained a predominantly immature phenotype, HLA-DR-medium,

CD8 O-I0 W, CD8 6 -negative and CD83-negative (Fig. 5.6A). Analysis of percentage CD 8 6 -positive

cells (Fig. 5.6B) and CD83-positive cells (Fig. 5.6C) clearly showed that a larger population of

UDP-glucose-treated cells express these maturation-associated markers than UDP-treated cells (n =

11, p = 0.001 for CD 8 6 comparison and p = 0.005 for CD83 comparison).

RNA was also extracted from these treated DC preparations and subjected to microarray profiling,

as detailed in Chapter 4. Analysis of the transcription of chemokines and chemokine receptors provided further evidence for UDP-glucose-mediated DC maturation (Fig. 5.7). As expected, DCs treated with TNF-a demonstrate a mature phenotype through elevated expression of CCL2, CCL17,

CCL22 and CCR7 transcripts, and reduced expression of CCL3, CCL4, CCL13, CCL15, CCL18,

CCL23, CCRl, CCR5 and GPRl05 transcripts, consistent with the expression pattern observed for

MCM matured DCs (Figs. 4.7 - 4.11). UDP-treated DCs showed similar transcript levels to untreated DCs, suggesting UDP has little effect on maturation. However, UDP-glucose-treated DCs

133 exhibited a trend towards the TNF-a-treated ‘mature DC’ gene expression profile. In comparison to both untreated and UDP-treated DCs, UDP-glucose-treated cells exhibited increased expression of

CCL2, CCL17, CCL22 and CCR7 transcripts, with a concomitant reduction in CCL3, CCL4,

CCL13, CCL15, CCL18, CCL23, CCRl, CCR5 and GPRl05 gene expression, consistent with induction of the DC maturation process.

An evaluation of the endocytic ability of UDP-glucose-treated DCs, as measured by FITC-dextran uptake, is shown in Figure 5.8. As expected, TNF-a-treated immature DCs demonstrate a decreased endocytic capability consistent with their differentiation towards maturity. The FITC-dextran uptake by immature DCs was unaffected by exposure to the negative control, UDP. Of interest,

UDP-glucose-treated cells showed a significant reduction in FITC-dextran uptake when compared to UDP-treated cells (n = 10, p = 0.0005). These data suggest that UDP-glucose, acting via the

GPRl 05 receptor, induces a DC maturation process characterised by reduced endocytic function.

134 B. UDP UDP-glucose TNF-a CD86

p = 0 . 0 0 1 iULLL 50n I CD80 LLL l i l l I CD86 I

C D 83 LfLfLL 50-1 p = 0.005 HLA-DR

CD83

N one UDP-glucose

UDP T N F -a

Figure 5.6 CD80, CD86, HLA-DR and CD83 expression on UDP-glucose-treated DCs

DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with or without UDP-glucose (ImM added to culture on day 6 and again on day 7), UDP (ImM added to culture on day 6 and again on day 7) or TNF-a (IpM on day 6 only) and the cells were then stained with fluorescently labelled antibodies to CD80, CD86, HLA-DR and CD83 for 30 minutes at 4°C and analysed by flow cytometry. A. Cell-surface expression of CD80, CD86, HLA-DR and CD83 in untreated (open histograms) or treated (closed histograms) DCs. One representative experiment of 11 is shown. B. Summary of cell-surface CD86 expression of untreated DCS, compared with UDP-, UDP-glucose- and TNF-a-treated DCs. C. Summary of cell- surface CD83 expression of untreated DCS, relative to UDP-, UDP-glucose- and TNF-a-treated DCs. Values represent the percentage of DCs expressing cell-surface CD86 (B) and CD83 (Q . Data shown are means (+SEM), n = 11.

135 CCR1 CCR5 CCR7

O 100 O 100 ü 400 Q)

CCL3 CCL4 CCL2

O 100 O 100 U 2000-,

03 1 0 0 0 -

CCL17 CCL13 CCL15

ü 100 O 75 O 500 û 50 03 250 25

CCL18 CCL23 CCL22

O 100 ü 100 ü 2000-1

2 50 03 50 2 1000-

GPR105

100 ü ITNF

03 lUDP 03 50 I UDP-glucose

Figure 5.7 The effect of UDP-glucose on transcript expression profiles in DCs

DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours (n = 2) or 72 hours (n == 2) with or without UDP-glucose (ImM added to culture on day 6, repeated daily), UDP (ImM added to culture on day 6, repeated daily) or TNF-a (IpM on day 6 only). RNA was extracted and reverse transcribed to ^^P-labelled cDNA, which was hybridised to a custom gene microarray. Values are expressed as the percentages of expression levels exhibited by untreated DCs, data are means (+SEM) of 4 experiments.

136 p = 0.0005

r- 10 0

u 50

Figure 5.8 The effect of UDP-glucose on the endocytic ability of DCs DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with or without UDP-glucose, UDP (ImM added to culture on day 6 and again on day 7) or TNF-a (IpM on day 6 only) and the cells were then incubated with FITC- dextran for 30 minutes at 37°C and fluorescence (indication of FITC-dextran uptake) was analysed by flow cytometry. Values are expressed as a percentage of the uptake by untreated DCs. Data shown are means (+SEM), n = 10.

137 5.2.5 The effect of endotoxin contamination on UDP-glucose-induced CD86 expression

Endotoxin has the potential to initiate most of the functional effects demonstrated in Section 5.2.4.

Therefore UDP and UDP-glucose, purchased from Sigma, were evaluated for endotoxin levels using the amebocyte limulus assay. Unfortunately low levels of endotoxin were present in both of the UDP-glucose preparations, see Table 5.2. With this in mind, experiments using the endotoxin binder and neutraliser, polymyxin B, were performed. The detected increase in CD86 expression correlated with the level of endotoxin contamination in the UDP-glucose preparations. Furthermore, polymyxin B abrogated the functional effects of UDP-glucose on CD86 upregulation in this experiment (Fig. 5.9). However, intracellular calcium data was not affected by endotoxin, as demonstrated by the observed intracellular calcium response to UDP-glucose in both the presence and absence of polymyxin B (Fig. 5.10).

Agonist Manufacturer and batch no. Endotoxin Units / ml

UDP-glucose Sigma, #7017 18.9 EU/ml

UDP-glucose Sigma, #7019 1.7EU/ml

UDP Sigma, # 7077 >0.1 EU/ml

Table 5.2. Endotoxin levels in UDP and UDP-glucose preparations Endotoxin levels in UDP and UDP-glucose preparations were determined using the amebocyte limulus assay.

138 no polymyxin B 1 0 0 -, CD CO + polymyxin B Q O 75-

0 50- ? O Q 25-

0 1 (O S c/3I c m03 N M- CL o 0 O O LL '■ a N, N 3 CL < 03 03 Q c o 3 03 1 Û. C Q Q 3 3

Figure 5.9 The effect of polymyxin B on UDP-glucose-induced CD86 expression by DCs DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours in the presence of UDP-glucose (UDP-glu), UDP, TNF-a or LPS. Batch numbers of UDP-glucose and UDP are indicated (also see Table 5.2). Polymyxin B was added where appropriate, at a final concentration of 1 mg/ml, four hours before addition of the agonists. Cells were then stained with a FITC labelled antibody to CD86 and analysed by flow cytometry. The percentages of DCs expressing CD86 are shown. Data are means + SEM of 3 experiments.

139 Histamine CCL3

20000-1 25000-1

20 0 0 0 - 1 5 0 0 0 -

5 0 0 0 - H 10000- LL ^ 10000 - 5 0 0 0 - 5 0 0 0 -

0 60 120 180 240 0 60 120 180 240

Time (seconds) Time (seconds)

UDP-glucose

6000-1 no polymyxin B ■H—- + polymyxin B i 3 0 0 0 -

0 60 120 180 240 Time (seconds)

Figure 5.10 The effect of polymyxin B on histamine, CCL3 and UDP-glucose-initiated intracellular calcium release in immature DCs DCs were derived from monocytes in the presence of GM-CSF and IL-4. Calcium responses were induced by histamine (lOOpM), CCL3 (IpM) and UDP-glucose (ImM, containing 3 EU/ml) in day 9 immature DCs +/- 4 hours pre-treatment with polymyxin B. Calcium respones were measured using the calcium sensitive dye Fluo-4-AM in conjuction with a FLIPR. Fluorescence intensity units (F.l.U.) are shown. Data are triplicates + SD.

140 5.2.6 Endotoxin free UDP-glucose

Endotoxin free UDP-glucose preparations were sourced from ICN Biomedicals (<0.01 EU/ml, assayed using the amebocyte limulus assay). Immature (day 6) DCs were cultured either with or without this endotoxin-free UDP-glucose and expression of CD86, often cited as a classical marker of DC maturation (Banchereau and Steinman, 1998; Banchereau et al., 2000), was quantified by immunofluorescent staining and flow cytometry. In general there was an increase in the percentage of CD86-positive cells in DCs treated with endotoxin-free UDP-glucose, although not all donors respond to this agonist (Fig. 5.11).

Since DCs derived from only certain donors were responsive to UDP-glucose, the possibility that

UDP-glucose acts in an additive or synergistic manrier with LPS to promote activation of DCs was investigated. All immature DC populations exhibited an increase in the percentage of cells expressing cell-surface CD86 in response to LPS treatment. The combination of UDP-glucose and

LPS resulted in a modest, but statistically significant, increase in CD86 expressing cells compared to LPS alone. Although LPS markedly increased CD83 expression and decreased FITC-dextran uptake in these cells, no statistically significant effect was seen with UDP-glucose for these functional parameters (Fig. 5.12). Allogeneic T cell proliferation experiments using these treated

DC preparations, and IFN-y measurements from the same experiments, show increased responses, although not statistically significant, in response to UDP-glucose treatment, both in the absence and presence of LPS (Fig 5.13).

Due to donor variability observed in the studies of CD86 upregulation, UDP-glucose-induced calcium fluxes and UDP-glucose-induced CD86 expression were evaluated in parallel in four donors. These data revealed a correlation between the magnitude of a UDP-glucose-induced calcium flux and an increase in CD86 expressing cells in these experiments (r^ = 0.87) (Fig. 5.14).

141 50-1

CO 00 [2 8 25- o Q

No treatment + UDP-glucose

Figure 5.11 The effect of endotoxin free UDP-glucose on CD86 expression by immature DCs DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with or without endotoxin-free UDP-glucose (ImM added to culture on day 6 and again on day 7), following which cells were stained with a FlTC-labelled anti-CD86 antibody and analysed by flow cytometry. The percentages of DCs expressing CD86 are shown. Individual data points represent individual donors; the median value is indicated.

142 A. B. CD86 CD83

p < 0.05 50

W 00 £ g 25 U) 00 O Q

[LPS] ng/ml [LPS] ng/ml C.

FITC-Dextran Uptake I I no UDP-glucose ■ ■ + UDP-glucose O 100-1

^ S 50 X 2 ° O Q à

[LPS] ng/ml

Figure 5.12 The effect of UDP-glucose and LPS on CD86 expression and CD83 expression and FITC-dextran uptake by DCs DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with or without UDP-glucose (ImM added to culture on day 6 and again on day 7) and LPS (1 ng/ml added day 6 only). Cells were then stained with fluorescently-labelled antibodies to CD86 (A) or CD83 (B) or incubated with FITC-dextran for 30 minutes at 37°C (C) and subsequently analysed by flow cytometry. Values are expressed as the percentage of positive DCs. Means + SEM are shown; n = 10 for CD83 and CD86 staining, n = 7 for FITC-dextran uptake.

143 IFN-y production

1 .0 0

0.75 D) 0.50

0.25

0.00 (D (/) (D U

Q 3 + COQ_

B. T cell proliferation

100000 Ü

E .9 50000

CO o

CO

Figure 5.13 The effect of UDP-glucose pre-treatment of DCs, with or without LPS, on allogeneic T cell proliferation and IFN-y production DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with UDP-glucose (ImM added to culture on day 6 and again on day 7); LPS (1 ng/ml added day 6 only); both UDP-glucose and LPS; or MCM (1ml on day 6 only). T cells were purified using the MACS magnetic bead system and a pan T cell kit. DCs were washed extensively, irradiated ('X s) and incubated with allogeneic T cells at a ratio of 1 DC to 100 T cells, in round-bottomed 96 well plates for 5 days. (A) Supernatant was removed on day 5 and IFN-y levels were assessed by ELISA. (B) Proliferation was measured by incorporation of thymidine over the final 16 hours of culture (irradiated DCs alone did not incorporate Thymidine or produce detectable levels of IFN-y (data not shown)). Data shown are mean + SEM of 3 donors.

144 10000-1

c

ccX 5000- D

LL

0 10 20

%CD86 +ve DC

Figure 5.14 Correlation of UDP-glucose-induced CD86 expression and UDP-glucose-induced calcium flux DCs were derived from monocytes in the presence of GM-CSF and IL-4. Day 6 DCs were incubated for 48 hours with or without UDP-glucose, stained with a fluorescently labelled antibody to CD86 and analysed by flow cytometry. Alternatively, UDP-glucose-induced calcium responses were measured using the calcium sensitive dye, Fluo4-AM, in conjunction with a FLIPR. The percentages of DCs expressing CD86 were plotted against the peak Ca^^ increases in fluorescence units minus basal (F.l.U. Max - Min). Points represent individual experiments; the linear regression is shown, r^= 0.87.

145 5.2.7 The effect of pertussis toxin pre-treatment on endotoxin-free UDP-glucose-induced

calcium flux in immature DC

Pertussis toxin is a well-established inhibitor of certain G proteins. To demonstrate the G-protein dependency of the UDP-glucose effect on DCs, the intracellular calcium responses to histamine,

CCL3 and UDP-glucose were investigated in immature DCs, with and without pertussis toxin pre­ treatment of the cells. The calcium flux of immature DCs to CCL3 was abolished by pertussis toxin pre-treatment (Fig. 5.15C) while the response to histamine remained intact (Fig. 5.15A). These data are consistent with the pertussis toxin-sensitivity of the Gi proteins known to couple to the CCL3 receptors, CCRl and CCR5 (Luther and Cyster, 2001) and the pertussis toxin resistant nature of the

Gq and Gs proteins that couple to the histamine receptors HI and H2 respectively (Hill et al., 1997), all of which are present and functional on monocyte-derived immature DCs (Mazzoni et ah, 2001).

The UDP-glucose-initiated calcium flux was down-modulated in response to pertussis toxin pre­ treatment (Fig. 5.15B). A paired t-test revealed statistically significant down-regulation of the UDP- glucose-initiated calcium flux by pertussis toxin pre-treatment for 10|iM UDP-glucose (n = 5, p =

0.02) and 100|iM UDP-glucose (n = 5, p = 0.003). This demonstrates that the increase in intracellular calcium in response to UDP-glucose is at least in part mediated via a pertussis toxin sensitive G protein.

146 A. B. Histam ine UDP-glucose

12000-1 5000-1

8000- ^ 2500- j 4000- u_

0.01 0.1 1 10 100 1000 0.1 1 10 100 1000 10000

[Ligand] |iM [Ligand] |aM

C. CCL3

12000-1 c —•— no pertussis toxin 8000- —0-- + pertussis toxin X

B 4000- u_

0.0001 0.001 0.01 0.1 1 10

[Ligand] pM

Figure 5.15 The effect of pertussis toxin pre-treatment on the calcium responses of immature DCs to histamine, CCL3 and endotoxin free UDP-glucose DCs were derived from monocytes in the presence of GM-CSF and IL-4, Where indicated, day 8 or 9 immature DCs were pre-treated with 10 ng/ml pertussis toxin for 18 hours following which A. Histamine, B. UDP-glucose and C. CCL3 induced calcium responses were measured using the calcium sensitive dye, Fluo4-AM, in conjunction with a FLIPR. Peak increases in fluorescence units minus basal (F.l.U. Max - Min) are shown. Data shown is mean +/- SEM of 5 donors. *p = 0.02, **p = 0.003, paired t-test.

147 5.2.8 Intracellular calcium release from GPRlOS-expressing HEK293 cells in response to

necrotic cell components

UDP-glucose plays a well-established role in metabolism and, as such, provides a potential candidate as a molecule for signalling the presence of damaged cells. The possibility that necrotic cells could release unusually high levels of UDP-glucose into the extracellular milieu is one such scenario. With this in mind, the release of intracellular calcium from GPRl 05 / Gai6-expressing

HEK293 cells in response to necrotic cell supernatants was investigated. Endothelin generated similar calcium flux data in both the GPRl 05 / Gai6-expressing HEK293 cells and the wild type

HEK293 cell line (Fig. 5.16). UDP-glucose, meanwhile, initiated a calcium flux only in the

GPRl 05 / Gtti6-expressing cell line. Necrotic cell supernatants, but not control, PBMC supernatants, induced a dose-dependant increase in intracellular calcium in GPRl 05 / Ga%6- expressing cells and not the wild type cell line (Fig. 5.16).

148 A. GPR105/Ga16 HEK293

20000-1 "2 10000-L

4 0 0 0 - r k 3 0 0 0 - zj 2000- LL 1000- 0-

10 3.3 1.1 .04 Neat 1in3 1in9 1in27

[Ligand] ^iM supernatant dilution

en d o th elin PBMC control supnt

UDP-glucose PBMC Necrotic supnt

B. HEK293

20000-1 c 10000-L

4 0 0 0 - r X (0 3 0 0 0 - zj 2000- i L 1000- 0-

—I------1------1------1— H I 1------1------i----- 10 3.3 1.1 .04 Neat 1in3 1in9 1in27

[Ligand] |iM supernatant dilution

Figure 5.16 GPRlOS-expressing HEK293 and control HEK293 responses to necrotic cell supernatants HEK293 cells were co-transfected with Gai6 and GPRl 05 and a stable cell line was obtained by serial dilution cloning. Calcium responses induced with endothelin, UDP-glucose, viable cell and necrotic cell supernatants were measured using the calcium sensitive dye, Fluo4-AM, in conjunction with a FLIPR; in GPRl 05 / Gai6-expressing HEK 293 (A) and HEK 293 wild type (B) cell lines. Peak increase in fluorescence units minus basal (F.l.U. Max - Min) are shown, data are triplicates +/- S.D. One representative experiment of 2 is shown.

149 5.3 Discussion

The most exciting observation from the microarray data detailed in Chapter 4 was the differential expression pattern of the GPR105 gene in monocytes and DCs. Immature DCs express GPR105 at substantially higher levels than mature DCs, with expression by monocytes similar to that of mature

DCs. Interestingly, significant ^pression of the GPR105 transcript has not been observed by any other cell type analysed with this microarray; including primary human kératinocytes and bronchial epithelial cells, in viïro-derived osteoclast cultures and primary T cells, as well as the cell lines

HUT-78 (T cell line) and Caco-2 (human intestinal cell line) all with and without treatment with various proinflammatory stimuli (data not shown). The expression of GPR105 in day 6 immature

DCs confirms that lack of expression in mature DCs is due to active down-modulation of gene transcription, while the expression by day 9 immature DCs is not solely a result of a lack of maturation signals in the later part of the incubation period. GPR105 transcript expression in day 6 and day 9 immature DCs, but not day 9 mature DCs, was confirmed by Northern blot analysis.

A stable cell line expressing GPR105 and the promiscuous G protein subunit, Gtti6, was used to confirm agonist capabilities of the putative GPR105 ligand, UDP-glucose (Chambers et al., 2000).

Intracellular calcium increases were seen in response to UDP-glucose in the GPR105 / Gai6- expressing cell line, and not in the parent HEK293 cell line. Calcium responses to other ligands, namely UDP (negative) and endothelin (positive), were equivalent in both the GPRl 05-expressing and the parent cell lines. UDP-glucose specific signalling was over two orders of magnitude more sensitive than UDP-induced signalling in the GPRl 05-expressing cell line. These data confirm that

UDP-glucose is an agonist for the GPRl 05 cell-surface 7TM molecule.

Following confirmation of the specificity of UDP-glucose for GPRl 05, the calcium signalling responses of monocytes, immature and mature DCs in response to UDP-glucose were evaluated.

Examining calcium fluxes in response to CCL3 and CCL19 controlled these experiments; the

150 receptors that respond to CCL3, CCRl and CCR5 present on monocytes and immature DCs, are downregulated on mature DCs, and CCR7, responsive to CCL19 is upregulated on mature DCs

(Sozzani et ah, 1998). As expected, monocytes and immature DCs both exhibited increased intracellular Ca^^ concentrations in response to CCL3 whereas mature DCs showed elevated intracellular Ca^^ levels in response to CCL19. UDP-glucose initiated a calcium flux in immature dendritic cells, but not mature DCs or monocytes. This correlates with the expression pattern of

GPRl 05 mRNA transcripts, adding support for the expression of functional GPRl 05 by the immature DC subset.

Specific signalling of UDP-glucose via GPRl 05 binding in immature DCs warranted further investigation of the role of this receptor in DC development. To this end, immature DCs were incubated with UDP-glucose and analysed extensively. In this series of experiments, TNF-a was used as a positive control for DC activation. Although TNF-a alone is inefficient at inducing fully mature DCs (Reddy et ah, 1997), it was employed here as a well defined positive control capable of inducing upregulation of CD80, CD83, CD86 and HLA-DR and decreasing the endocytic ability of immature DCs (Sallusto and Lanzavecchia, 1994). Flow cytometric analyses revealed increased cell-surface expression of CD80, CD83, CD86 and HLA-DR following stimulation with TNF-a and a similar phenotypic change following stimulation with UDP-glucose, by way of an upregulation in

CD80, CD83, CD86 and HLA-DR expression. In particular, significant upregulation of CD83 and

CD86 was seen when compared to the non-GPR105 binding UDP control. Analysis of mRNA transcript expression in UDP-glucose treated DCs revealed a pattern consistent with both TNF-a- treated DCs and the MCM-matured DCs analysed in Chapter 4. Increases in transcripts encoding

CCL2, CCL17, CCL22 and CCR7, with a reduction in CCL3, CCL4, CCL13, CCL15, CCL18,

CCL23, CCRl, CCR5 and GPRl05 were evident. In addition to the many changes in gene expression that occur during DC maturation, downregulation of the antigen internalisation capability of DCs is also observed. As antigen uptake can take place via endocytosis (Sallusto and

151 Lanzavecchia, 1994), the endocytic capacity of DCs was evaluated by measuring the uptake of

FITC-dextran, in order to determine if the apparent UDP-glucose-induced shift towards a mature phenotype affected this ftmctional marker of DC development. As expected, TNF-a-treated DCs exhibited a decrease in endocytic ability consistent with their progression towards maturity, whereas the endocytic activity of immature DCs was unaffected by exposure to UDP. UDP- glucose-treated immature DCs, meanwhile, also showed a significant reduction in FITC-dextran uptake when compared to UDP-treated cells. These data support the hypothesis that UDP-glucose induces a mature DC phenotype, both in terms of cell-surface molecule expression and ftmctionally.

Endotoxin is a lipopolysaccharide (EPS), which is a component of the outer membrane of gram- negative bacteria. Endotoxin or EPS has long been identified as an initiator of DC maturation

(Verhasselt et al., 1997). It was therefore essential to eliminate possible endotoxin contamination of the UDP-glucose preparations used in these experiments. Unfortunately, the amebocyte limulus assay revealed low level endotoxin contamination of the UDP-glucose preparations. Further experiments were performed in the presence of the endotoxin binder and neutraliser, polymyxin B.

CD86 induction by UDP-glucose correlated with the extent of endotoxin contamination in the UDP- glucose preparations. Furthermore the induction of CD86 expression was abrogated by the presence of polymyxin B. These data suggest that the increase in CD86 initiated by UDP-glucose is, at least in part, due to endotoxin contamination of the UDP-glucose preparations. However, UDP-glucose- induced intracellular calcium fluxes were not affected by the addition of polymyxin B. Indeed, endotoxin, signalling through the Toll-like receptors in an EPS activation cluster, would not be expected to induce a calcium signal (Triantafilou and Triantafilou, 2002).

Thereafter, endotoxin-free UDP-glucose preparations were evaluated for their ability to induce

CD86 in immature DCs. In some donors, UDP-glucose-treated DCs were induced to express increased levels of the co-stimulatory molecule CD86. The relatively low level of this response suggested that thresholds for activation of the DCs were not effectively reached. Extracellular

152 nucleotides, such as ATP, synergise with other molecules to induce dendritic cell maturation

(Schnurr et ah, 2000), furthermore extracellular nucleotides can modulate the actions of LPS

(Tonetti et ah, 1995; Ferrari et ah, 1997; Proctor et ah, 1994; Ferrari et ah, 2000; Denlinger et ah,

2001). Therefore the higher level of CD86 expression observed with the endotoxin contaminated

UDP-glucose preparations, prompted the investigation of a possible synergy between UDP-glucose

and LPS. LPS at Ing/ml initiated increased CD86 and CD83 expression and decreased FITC-

dextran uptake in immature DCs. However, these effects were not of the magnitude seen with monocyte-conditioned medium demonstrated in Chapter 3, allowing for potential additive or

synergistic effects of UDP-glucose to be evaluated. Using a combination of LPS and UDP-glucose

to induce CD86 expression resulted in a statistically significant increase over the levels of CD86

seen with LPS alone. However, CD83 expression and FITC-dextran uptake were unaffected by the

addition of UDP-glucose, either in the presence or absence of LPS. An evaluation of the ability of

UDP-glucose-treated immature DCs to stimulate naive T cells demonstrated an increase over

untreated DCs. A slight additive effect with LPS was also evident. These differences are not

statistically significant and could be considered negligible; however a trend is certainly evident.

Furthermore, in four donors the induction of CD86 by endotoxin-free UDP-glucose preparations

was compared to the magnitude of an UDP-glucose-induced calcium flux in these same donors.

This revealed a very strong correlation between the calcium flux levels and the increase in CD86

expression. This suggests that where intracellular calcium release is low, UDP-glucose-induced

signalling has not reached the magnitude required for the induction of dendritic cell maturation.

To further substantiate that UDP-glucose functions through GPRl 05 ligation, the pertussis toxin

sensitivity of an UDP-glucose-initiated calcium flux was investigated in immature DCs. The ability

to block a calcium flux with pertussis toxin is limited to receptors coupled with pertussis toxin-

sensitive Gi G-proteins. A previous report, using transfected cell lines, suggests that GPRl 05 is

coupled to Gi G-proteins (Chambers et al., 2000). The calcium flux of immature DCs to CCL3 was

153 abolished with pertussis toxin pre-treatment and the response to histamine remained intact. These data are consistent with the pertussis toxin-sensitive nature of the Gi proteins known to couple the

CCL3 receptors, CCRl and CCR5 (Luther and Cyster, 2001) and the pertussis toxin-resistant nature of the Gq and Gs proteins that associate with the histamine receptors HI and H2 respectively (Hill et al., 1997), present and functional on monocyte-derived immature DCs (Mazzoni et al., 2001; la

Sala et al., 2002; Idzko et al., 2002; Gutzmer et al., 2002). The UDP-glucose-initiated calcium flux via GPRl05 was downmodulated in response to pertussis toxin pre-treatment of immature DCs; however the signal was not abolished. This demonstrates that the calcium signal obtained in response to UDP-glucose is, at least in part, mediated by pertussis toxin-sensitive G proteins. There is precedence for GPCRs to associate with different G protein classes and therefore display incomplete sensitivity to pertussis toxin (Haribabu et al., 1999; Ahamed and All, 2002).

What is the physiological consequence of UDP-glucose-mediated DC signalling? Although the downstream effect of this UDP-glucose initiated calcium flux is not clear, the presence of UDP- glucose and other glycolytic metabolites at high levels in tumour cells (Eigenbrodt et ah, 1992) is intriguing. This leads to speculation that UDP-glucose could be acting as a tumour-specific ‘danger’ signal to the immune system. Indeed, supernatants from necrotic cancer cells, but not other cell types, can induce maturation events in immature DCs (Sauter et al, 2000). Furthermore, immunogenic tumours, such as melanomas (Brinckerhoff et ah, 2000), are present in microenvironments where immature DCs, such as Langerhans cells, are highly prevalent. Perhaps increased levels of UDP-glucose released from necrotic tumour cells might reach a ‘danger’ threshold and, via GPRl 05-mediated signalling, in concert with other signals such as LPS, initiate an immune response, following induction of DC maturation. Furthermore, UDP-glucose has a well- established role in glycolytic metabolism and, until recently, had not been detected extracellularly.

This leads to the possibility that UDP-glucose could be released inappropriately from damaged or dying cells. With this in mind, the agonist activities of necrotic cell supernatants were evaluated

154 using GPRl 05-expressing HEK293 cells. Indeed, supernatants from necrotic PBMC induced a calcium response in GPRl 05-expressing cells, whereas no response was seen in the parent HEK293 cell line. However, the release of UDP-glucose from resting cells has also recently been demonstrated (Lazarowski et al., 2003). Lazarowski and colleagues demonstrate that unlike other nucleotide signalling molecules such as ATP, release of which following mechanical stimulation of

cells reaches a peak at ten minutes and diminishes rapidly, UDP-glucose release is initially relatively minor. However, levels of UDP-glucose remain stable for several hours, suggesting that

the rate of metabolism of extracellular UDP-glucose is much lower than that of ATP. These data

add credence to the suggestion that a critical threshold of UDP-glucose may have to be reached for

GPRl 05 signalling to be initiated.

Recently Lee et al. have reported GPRl 05 expression identifies a subset of haemopoietic cells

within foetal bone marrow (Lee et al., 2003). They also demonstrate chemotaxis of these cells to a-

conditioned medium from bone marrow stroma and that this migration is specifically blocked by an

anti-GPR105 antibody, implicating GPRl 05 in the homing of bone marrow stem cells. The bone

marrow microenvironment is a rich source of immature dendritic cells and it is possible that the

bone marrow stem cell subset described by Lee et al. is a precursor population for DCs.

Interestingly, the authors also describe more efficient responses of GPRl 05-transduced human

CD34-positive cells to bone marrow stroma-conditioned media than UDP-glucose and suggest the

possibility of a distinct ligand.

The work described in this chapter demonstrates GPRl 05 expression on immature monocyte-

derived DCs and its absence from mature DCs and monocyte precursors. Initial evaluation of the

functional effects of UDP-glucose on immature DCs was complicated by the presence of low levels

of endotoxin in the UDP-glucose preparations. However, initiation of DC maturation by endotoxin

free UDP-glucose was achieved in some donors. In response to UDP-glucose intracellular calcium

increased in immature DCs, but not mature DCs and monocytes. These data were confirmed using

155 endotoxin-free UDP-glucose. The increase in intracellular calcium induced by UDP-glucose demonstrated partial sensitivity to pertussis toxin, consistent with the previous description of this receptor as a Gi linked receptor (Chambers et al., 2000). Finally, necrotic cell supernatants initiated an increase in intracellular calcium in a GPR105-expressing cell line and not the non-transfected parent cell line. These data demonstrate signalling by UDP-glucose in immature DCs, correlating with the expression profile of GPRl 05. Possible downstream effects of this UDP-glucose-mediated signal are suggested, although an unequivocal role remains elusive at this time.

Further evaluation of the significance of GPRl 05 expression on immature DCs would require demonstration of GPRl 05 on purified blood DCs. Although extensively used as tools for investigating myeloid DCs, representation of the true in vivo myeloid population by in vitro monocyte-derived DCs is still open to question. Definitive confirmation that UDP-glucose is acting through GPRl 05 on immature DCs awaits verification with either specific anti-GPR105 antibodies or antagonising chemical entities or other specific gene targeting approaches. Therefore the tentative role we describe for UDP-glucose-induced signalling in DC maturation requires further confirmation with better tools than were available for these studies.

156 Chapter 6

CXCL5 in osteoclasts

157 Chapter 6

6.1 Introduction

Osteoclasts are specialised cells of haematopoietic origin, belonging to the myeloid cell lineage.

The importance of chemokines and chemokine receptors in the biology of other myeloid cell types is well documented. However, despite some emerging information little is known of the significance of these molecules in osteoclast physiology and function. With this in mind, the custom made gene microarray was used to examine expression of mRNA transcripts encoding chemokines and chemokine receptors, as well as additional selected GPCRs, in human monocytes and in vitro monocyte-derived osteoclasts as described in Chapter 4. The resulting gene expression data revealed much similarity of expression of these genes by the two cell types, perhaps reflecting the relative insignificance of these molecules in osteoclast specialisation. However, some genes did show distinct expression patterns as previously discussed in Chapter 4, most interestingly CXCL5

(ENA-78), an ELR motif-containing CXC chemokine implicated in neutrophil recruitment and angiogenesis, showed increased expression in the osteoclast samples.

CXCL5 also called epithelial neutrophil activating peptide 78 (ENA-78) or lipopolysaccharide- induced CXC chemokine (LIX), is a potent chemoattractant and activator of neutrophil functions. It shares a receptor, CXCR2, and neutrophil activating properties with CXCL8 (IL-8) (Ahuja et al.,

1996). CXCL5 is produced at low levels by many cell types and is elevated in the inflamed tissues of patients with rheumatoid arthritis (RA); Crohn's disease; ulcerative colitis; acute appendicitis and allergic airway inflammation (Walz et al., 1997). CXCL5 is also a potent angiogenic factor in non small cell lung cancer and attracts endothelial cells with the same influence as CXCL8 (Arenberg et al., 1998; Strieter et al., 1995).

158 The CXC chemokine family can be subdivided by the presence of the tri-amino acid motif Glu-Leu-

Arg (the ELR motif): CXCL5 and CXCL8 form part of the ELR-positive CXC chemokine subfamily, along with CXCLl (Gro-a), CXCL2 (Gro-(3), CXCL3 (Gro-y), CXCL6 (GCP2) and

CXCL7 (NAP-2). This motif affords activity through the chemokine receptors CXCRl and

CXCR2, although the ELR motif containing chemokines display differing abilities to agonise these receptors (Clark-Lewis et al., 1993; Baggiolini et al., 1994). Chemokines that do not contain the

ELR motif, such as CXCL4 (PF4), CXCL9 (MIG) and CXCLIO (IP 10) do not bind these receptors

(Clark-Lewis et al., 1993). Furthermore, characterisation of these CXC chemokines as angiogenic or angiostatic has been linked to the presence or absence of the ELR motif; family members that contain the ELR motif are potent promoters of angiogenesis, whereas members that lack this motif are inhibitors of angiogenesis (Arenberg et al., 1998; Belperio et al., 2000). CXCL8, CXCLl,

CXCL5 and CXCL6 are all capable of inducing endothelial cell chemotaxis in vitro and promoting angiogenic activity in the rat comeal micropocket model of neovascularisation (Strieter et al.,

1995). Conversely CXCL4 demonstrates anti-angiogenic affects both in vitro and in vivo and

CXCL4, CXCL9 and CXCLIO have been shown to potently inhibit angiogenesis induced by both

ELR-positive CXC chemokines and basic fibroblast growth factor (bFGF) (Maione et ah, 1990;

Strieter et ah, 1995)

The role of CXCL5 and CXCL8 in angiogenesis is particularly interesting with respect to the biology of osteoclasts; not least because of the intimate relationship between angiogenesis and bone resorption, both at the site of normal physiologic bone turnover and in pathological conditions such as rheumatoid arthritis (Paleolog, 1996). Indeed Vascular Endothelial Growth Factor (VEGF), an important inducer of angiogenesis and endothelial cell proliferation (Veikkola and Alitalo, 1999) has direct effects on osteoclast formation and function (Nakagawa et ah, 2000; Collin-Osdoby,

1994) and can substitute for M-CSF in the support of osteoclastic bone resorption (Niida et ah,

1999). Consequently human microvascular endothelial cells (HMVEC) have been shown to express

159 RANKL, while HMVEC treated with TNF-a can support the formation and fusion of osteoclasts from human monocyte precursors with concomitant bone resorption by these in v/7ro-derived osteoclasts (Collin-Osdoby et ah, 2001; McGowan et ah, 2001). Furthermore, a hypoxic environment, known to be an inducer of the angiogenic mediator VEGF (Neufeld et ah, 1999), has recently been shown to enhance both osteoclast formation and bone resorption in vitro (Amett et ah,

2003). It is an attractive hypothesis that actively resorbing osteoclasts at the site of bone turnover may secrete CXCL5 and CXCL8 in response to angiogenic signals, thereby potentiating endothelial cell recruitment and new blood vessel formation. In addition, differential receptor usage by CXCL5 and CXCL8 adds the possibility of disparate roles for these chemokines.

High levels of CXCL8 production have already been attributed to osteoclasts in vitro and CXCL8 mRNA has been detected in osteoclasts present in RA joints by in situ hybridisation (Rothe et ah,

1998). This chapter aims to demonstrate expression of CXCL5 and CXCL8 proteins by osteoclast cultures and to attribute the expression to multinucleated cells by immunocytochemistry. The functionality of CXCL5 secreted by osteoclasts will be confirmed by induction of neutrophil chemotaxis. Lastly, the influence of inflammatory and angiogenic mediators on CXCL5 and

CXCL8 production by osteoclasts will be evaluated.

160 6.2 Results

6.2.1 Transcription and translation of CXC chemokine genes in osteoclast cultures

Four ELR-positive CXC chemokine transcripts were significantly expressed (>20 arbitrary units) in osteoclasts, as revealed by gene microarray profiling (Fig. 6.1 and Appendix 1). These included

CXCLl (Gro-a), CXCL7 (NAP-2) and CXCL8 (IL-8), each of which was also transcribed in monocytes. CXCL5 (ENA-78) demonstrates much higher mRNA expression in the osteoclast preparations than the monocytes (Table 4.1; Fig. 4.13; Fig. 6.1). Interestingly, none of the four myeloid cell types tested had significant expression of the ELR-negative chemokine genes CXCL9

- CXCLl4. However, the ELR-negative angiostatic chemokine CXCL4 was expressed at low levels in both osteoclasts and monocytes.

To confirm expression of these chemokines at the protein level, time course studies of in vitro osteoclastogenesis were performed and supernatants assayed using specific ELISAs. Since CXCL5 transcripts were not present in monocytes, cultures were set up with and without RANKL, which is necessary for multinucleated cell formation (Fuller et al., 1998), to help assess the relationship between CXCL5 production and osteoclast formation. Osteoclast cultures generated from monocytes of three donors were evaluated for production of CXCL5 as well as CXCLl, CXCL7,

CXCL8, CXCL9 and CXCLIO over a period of 9 days. On day 9, cultures were assessed for formation of TRAP-positive multinucleated cells. RANKL positive cultures demonstrated an average of 134 +/- 32.8 (SEM) TRAP-positive multinucleated cells per field of view. No multinucleated cell formation was detected in the absence of RANKL. Figure 6.2 reveals production of CXCLl and CXCL7 at similar levels in both the presence and absence of RANKL.

CXCL8 levels, although higher in the later time points of the RANKL cultures, were similar in both cultures at the early time points. Production of CXCL5 appeared to be dependent on the presence of

RANKL in the cultures and increased dramatically on days 5 and 6 of culture. ELR-negative

161 chemokines CXCL9 and CXCLIO were mostly below the limit of detection of the ELISAs. Very low levels of CXCLIO could be measured at day 9 in both cultures.

1 5 0 - £ M on ocytes II Osteoclasts o. 1 0 0 - g SI a: CO 5 0 -

.M _ y 1 LO 00 —I _i Ü o o o o o g g s g g g

Figure 6.1 Expression of CXCLl, CXCL5, CXCL7, CXCL8, CXCL9 and CXCLIO mRNA transcripts in monocytes and osteoclasts

Monocytes were purified from PBMC and osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridised to a gene microarray. Expression levels were calculated by normalising data to a panel of forty housekeeping genes. Data are means (+SEM) of6 (monocytes) or 7 (osteoclasts) experiments.

162 CXCL8 / IL-8 CXCL5 / ENA-78

50-1 7 5 -

O) D) c

00 [ n _l 2 5 -

§ 2 , 2 5 - IÉI t o J a 0 2 4 5 6 7 9 0 2 4 5 6 7 9 Time (days) Time (days)

CXCL1 /GRO a CXCL7 / NAP-2

2n

O) c § 1 2 4 5 6 7 9 0 2 4 5 6 7 Time (days) Time (days)

CXCL9 / MIG CXCL10/IP10

2n

O) O) c c

§ g ■n in in in in in in in in i i 4 5 6 7 0 2 4 5 6 7 Time (days) Time (days)

Figure 6.2 Levels of CXC chemokines in the supernatants of osteoclast cultures with or without RANKL

Osteoclast cultures, derived from PBM, were established in the presence of M-CSF with (closed

bars) or without (open bars) RANKL for 9 days. Supernatant was removed on days 0, 2, 4, 5,6 , 7 and 9. ELISAs were used to measure the CXC chemokines present in the supernatants, in duplicate. Data are means + SEM of 3 donors.

163 6.2.2 Immunocytochemical analysis of osteoclast chemokine production

As osteoclast cultures are not homogeneous, immunocytochemistry was applied to in vitro differentiation cultures on culture slides to identify the cell type responsible for production of

CXCL5. Day 9 osteoclast cultures were treated with GolgiStop™, which contains the protein transport inhibitor monensin, for 4 hours to allow the chemokine to accumulate in the cytoplasm of cells. Slides were fixed in a paraformaldehyde solution, permeabilised and blocked, before incubation with either goat IgG control, goat anti-CXCL5 or goat anti-CXCL8. Figure 6.3 demonstrates both CXCL5 and CXCL8 staining in multinucleated osteoclast cells. Staining of some of the mononuclear cells can also be seen in both cases, but that in multinucleated cells was most pronounced. No positive staining could be seen with the isotype control antibody.

164 $ # A- * •* Ô % *# *1SJ * **' # # • Control IgGl

% % I m , 1 * - . %# 4 * V ^ wL

B. 4 P

* • ' * ■ t f } * g * . * *S l * " •% - i t # ' ♦ 0 ’ # f anti-CXCL5 %

* I • «* $%

C.

#» ^ ■ I \ % *'V V « anti-CXCL8 : # * c #

Figure 6.3 CXCL5 and CXCL8 production in osteoclast cultures Osteoclast differentiation cultures were grown on 6-well culture slides (Falcon) for 9 days, in the presence of GolgiStop™ (4pl/ml) for the last 3 hours of culture. Slides were fixed in PBS/ 3.5% paraformaldehyde/ 2% sucrose, permeabilised in 20mM Hepes/ 300mM sucrose/ 50mM NaCl, 3mM M gC y 0.5% triton-xlOO, and blocked in PBS/ 5% FCS/ 2% sucrose, before incubation with lOpg/ml goat IgG control (A), lOpg/ml goat anti-human CXCL5 {B) or lOpg/ml goat anti-human CXCL8 (Q. Bound antibody was revealed using HRP conjugated mouse anti-goat Ig and AEG substrate. Slides were then counterstained with Haematoxylin Gill no. 3, mounted and pictures were taken using a Zeiss axioscope with a JVC 3 chip colour camera (magnification x200).

165 6.2.3 Chemotaxis to osteoclast supernatants

In an attempt to elucidate a role for CXCL5 production by osteoclasts, I sought to examine the potential pro-angiogenic activity of osteoclast supernatants. To achieve this, endothelial cell chemotaxis assays were set up using human embryonic kidney endothelial cells (HUVEC).

Unfortunately the ability of HUVEC to respond to serum-containing medium ruled out the possibility of investigating chemotaxis to osteoclast supernatants, which routinely contain 15 % serum (Fig. 6.4).

As CXCL5 and CXCL8 are both capable of attracting neutrophils, experiments were performed to demonstrate the functionality of the chemokines produced in these osteoclast cultures using neutrophil chemotaxis as a read-out. Standard curves of neutrophil chemotaxis to recombinant human CXCL5 and CXCL8 are shown in Figure 6.5. CXCL8 is a much more powerful chemoattractant than CXCL5 in this system. Thereafter, chemotaxis of neutrophils to the osteoclast supernatants was investigated, using neutralising antibodies to CXCL5 and CXCL8 to attribute any chemotaxis seen to these chemokines. Osteoclast supernatants from three individual donors all showed powerful chemoattractant properties that could be significantly blocked by pre-treatment with an antibody to CXCL8. The probability that decreased chemotaxis to the osteoclast samples, in the presence of blocking antibodies, is due to random events is < 0.05 as determined by paired t-test analysis. Treatment with an antibody to CXCL5 resulted in statistically significant blocking of chemotaxis to one osteoclast supernatant and the CXCL5 standard. While chemotaxis to the other two osteoclast supernatants was not significantly blocked, a decrease in the percentage of cells migrating to these antibody treated supernatants was apparent (Fig. 6.6).

166 50000- Diluent:

40000 EBM + 0.1% BSA EGM + 0.1% BSA 30000 * Il EGM + 2% FBS 20000

o 00 3il 10000 -

0- 10% sérum Diluent Chemoattractant

Figure 6.4 Endothelial cell chemotaxis to 10% serum HUVECs were labelled with Calcein AM and placed in the top chamber of 96 well fluoroblok plates. Chemoattractants (10% serum or diluent as a control) were then place in the bottom wells using the sample ports. The plates were incubated for 4 hours at 37°C, 5% CO 2 and migrated cells were estimated by measuring the fluorescence intensity of the bottom chamber using an LJL analyst and bottom read optics at 485nm (excitation) and 530nm (emission). Data shown are mean of duplicates. The error bars represent minimum and maximum values.

150-, CXCL5

CXCL8

o ^ 100- O (Di > c o 5 0 -

1 10 100 1000

[Chemokine] ng/ml

Figure 6.5 Chemotaxis of human peripheral blood neutrophils in response to CXCL5 and CXCL8 Neutrophils were purified from PBMC. Chemokines were added to the bottom of 3pm transwell chemotaxis chambers, and neutrophils were placed in the top chambers. The plates were incubated for 60 minutes at 37°C, 5% CO 2, following which the top wells were discarded before migrated cells were recovered by vigorous pipetting and counted on a flow cytometer. Data is expressed as the percentage of neutrophils that migrated to the positive control, fMLP. Data is mean +/- SEM of 3 separate experiments.

167 A. Anti-CXCL5

100n

fi 5 0 - « 0

Anti-CXCL8

B. 100-1

o CL Ü _1 ^ 5 0 - c o (2

CN 00 00 CLCL Q. T3

O) O) C c m If) c\j CN

Figure 6.6 Blocking of neutrophil chemotaxis to osteoclast supernatants by anti-CXCL5 and anti-CXCL8 antibodies Neutrophils were purified from PBMC. Chemokines or osteoclast supernatants, plus anti-CXCL5 (A) or anti-CXCL8 (B) where appropriate, were placed in the bottom of 3pm transwell chemotaxis chambers and pre-incubated at 37°C, 5% CO2, for 1 hour. Neutrophils were added to the top chambers. The plates were incubated for a further 60 minutes at 37°C, 5% CO 2 , following which the top wells were discarded and migrated cells were removed from the bottom wells by vigorous pipetting and counted on a flow cytometer. Neutrophil chemotaxis to osteoclast supernatants or controls was measured in the presence (solid bars) or absence (open bars) of (A) anti-CXCL5 and (B) anti-CXCL8. Data is mean + SEM of 3 separate experiments; * p < 0.05, paired t test.

168 6.2.4 Induction of CXCL5 and CXCL8 production by angiogenic and inflammatory

mediators

To investigate whether osteoclasts produce increased levels of CXCL5 in response to angiogenic stimuli, the effect of VEGF on these cultures was examined. No increase in the levels of CXCL5 or indeed CXCL8, were seen in response to VEGF (Fig. 6.7). TRAP-positive multinucleated cell numbers were similar in wells with or without VEGF (data not shown). To determine whether there could be a significant role for osteoclast production of CXCL5 in inflammation, chemokine levels were measured in response to the inflammatory stimuli TNF-a and IE-1(3. Both inflammatory stimuli caused an increase in CXCL5 and CXCL8 levels (Fig. 6.8). Consequently, a two-fold increase in the number of multinucleated cells in these cultures was also seen (data not shown).

(D 200 - c o c o o [— n CXCL5 3 -5 T3 100- 2 r — 1 CXCL8 CD a.

110 100 [VEGF] ng/ml

Figure 6.7 The effect of VEGF on CXCL5 and CXCL8 production from osteoclast cultures Monocytes were purified from PBMC. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 8 days. VEGF was added to day 6 osteoclast cultures. The cells were then incubated for a further 48 hours, following which supernatants were removed and chemokine levels measured by ELISA. Data is expressed as a percentage of the chemokine levels produced in the absence of VEGF. Data shown are mean of two independent experiments. The error bars represent minimum and maximum values.

169 c 2 0 -1 o

îl 1 ^ — I CXCL5 il 10- [— n CXCL8 ü o

IL -lp + TNF- a

Figure 6.8 The induction of CXCL5 and CXCL8 from osteoclast cultures by inflammatory mediators Monocytes were purified from PBMC. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. lOng/ml IL-1(3 or lOng/ml TNF-a was then added to day 6 osteoclast cultures. The cells were incubated for a further 72 hours, following which supernatants were removed and chemokine levels measured by ELISA. Data is expressed as fold induction of chemokine production in the absence IL-1(3 or TNF-a. Data shown are mean of duplicates. The error bars represent minimum and maximum values.

170 6.3 Discussion

The microarray data detailed in Chapter 4, revealed expression of the ELR motif-containing CXC chemokine genes CXCLl, CXCL5, CXCL7 and CXCL8 in osteoclast cultures. In particular the expression of CXCL5 was of interest as this appeared to be elevated in the osteoclast samples (Fig.

6.1). Expression of mRNA encoding CXCLl, CXCL5, CXCL7 and CXCL8 in the osteoclast cultures was substantiated by detection of the corresponding proteins in osteoclast culture supernatants. Time course experiments further demonstrated a correlation between levels of CXCL5 protein and the formation of TRAP-positive multinucleated osteoclasts. TRAP-positive multinucleated cells were shown previously to appear at day 6 of culture, with numbers rising sharply between day 7 and day 8 (Fig. 3.9). The CXC chemokines, CXCLl, CXCL7 and CXCL8 were all present in RANKL-independent cultures, in which no multinucleated cell formation was seen, and at early time points in the RANKL containing cultures, thereby failing to demonstrate a clear correlation with the formation of multinucleated cells. The non-ELR containing angiostatic chemokines CXCL9 and CXCLIO were not detected in the osteoclast cultures, aside from very low levels of CXCLIO at day 9 of culture. Although CXCL5 levels were undetectable in the cultures stimulated with M-CSF only, ruling out the possibility that CXCL5 production could be attributed to macrophage-like cells contaminating the RANKL-positive cultures, the possibility of RANKL induction of CXCL5 from monocytes remains a possibility. Therefore immunocytochemistry was performed on the osteoclast cultures to confirm the cell types associated with the production of the

CXC chemokines, CXCL5 and CXCL8. This analysis clearly attributed CXCL5 and CXCL8 production in the osteoclast cultures to multinucleated osteoclast cells. Although some synthesis of these chemokines in mononuclear cells was also evident, many of these mononuclear cells had an appearance that could be associated with the TRAP-positive osteoclast precursor cells seen in osteoclast differentiation cultures.

171 To examine the functionality of CXCL5 in the osteoclast supernatants, endothelial cell chemotaxis

was investigated. Endothelial cell chemotaxis as a readout for CXCL5 activity is preferable to

neutrophil chemotaxis, since CXCL5 and CXCL8 share equivalent activities with respect to

endothelial cell chemotaxis (Strieter et ah, 1995) whereas CXCL5 demonstrates a much reduced

ability to induce neutrophil chemotaxis than CXCL8 (Clark-Lewis et ah, 1993; Baggiolini et ah,

1994). Unfortunately a potent inducer of endothelial cell chemotaxis is serum (Harvey et ah, 2002)

and as the osteoclast supernatants contain a high level of serum, the ability of endothelial cells to migrate to serum was first evaluated. As expected the large response to serum eliminated the use of

this cell type in evaluating the chemotactic function of CXCL5 protein in the osteoclast

supernatants and the small amounts of supernatant obtained from these cultures made the removal

of serum proteins through dialysis or chromatography unfeasible.

CXCL8 demonstrated greater induction of neutrophil chemotaxis than CXCL5 (Fig.6.5) (Clark-

Lewis et ah, 1993; Baggiolini et ah, 1994). A neutralising antibody to CXCL8 could significantly block the osteoclast supernatant-induced chemotaxis of neutrophils. In contrast, although an

antibody to CXCL5 decreased the osteoclast supernatant-mediated chemotaxis significantly in only one donor, the number of cells migrating to the osteoclast supernatants decreased in all three cases

studied. Overall, functionality of CXCL5 and CXCL8 in osteoclast supernatants was demonstrated in these assays.

The correlation between osteoclast formation and CXCL5 production suggests a distinct role for

CXCL5 in osteoclast biology. CXCL5 is a potent chemoattractant and activator of neutrophil functions and shares neutrophil activating properties with CXCL8 (Ahuja et ah, 1996). However

CXCL8 appears to have a much more significant role in the assays described in this chapter (Clark-

Lewis et ah, 1993; Baggiolini et ah, 1994), whereas both CXCL5 and CXCL8 are equipotent in their ability to attract endothelial cells (Strieter et ah, 1995). Given that no ELR-negative CXC chemokines could be detected in osteoclasts (Chapter 3), and since there are many links between

172 osteoclast formation, activity and angiogenesis, the angiostatic nature of these ELR-negative chemokines versus the angiogenic nature of CXCL8 and CXCL5 is intriguing. The angiogenic activities of CXCL8 on human intestinal microvascular endothelial cells are mediated through

CXCR2 (Heidemann et al., 2003), which is the receptor of choice for CXCL5 (Ahuja et ah, 1996), suggesting an element of redundancy for CXCL5 and CXCL8 in this response.

Considering these differential activities of CXCL5 on neutrophils and endothelial cells, it is attractive to suggest that CXCL5 production is, despite an element of redundancy, a specific mechanism for osteoclast induced angiogenesis at the site of new bone formation. Nakagawa et al have reported enhancement of osteoclastic activity by the angiogenic factor VEGF, mediated through its receptors KDR and Flt-1 (Nakagawa et ah, 2000). With this in mind, the effect of VEGF on CXCL5 production by osteoclasts was evaluated. VEGF does not increase CXCL5 levels in osteoclast cultures, however this does not rule out an active angiogenic role for CXCL5 produced by osteoclasts.

The effects of inflammatory stimuli on CXCL5 levels in osteoclast cultures were also investigated.

Increased production of both CXCL5 and CXCL8 was detected when TNF-a or IE-1(3 were added to the cultures. However, increased numbers of multinucleated cell formation was also found in these cultures (data not shown); therefore it is possible that the increased chemokine levels were simply due to an increase in multinucleated cell numbers. Furthermore, the uncharacterised mononuclear contaminants within these cultures are very likely to be macrophage type cells

(Chapter 3) and an increase in production of CXCL5 and CXCL8 in these cells, at least in response to TNF-a, would be expected (Ciesielski et ah, 2002). Therefore, to confirm a possible correlation between CXCL5 production by osteoclasts and inflammatory disease states, a demonstration of

CXCL5 in osteoclasts at sites of inflammation would be required. CXCL8 has already been demonstrated in osteoclasts in the rheumatoid arthritis joint and indeed, the intimate relationship between bone resorption and angiogenesis is also found in rheumatoid arthritis where bone

173 destruction and invasion of blood vessels appear simultaneously (Paleolog, 1996). The data reported in this chapter demonstrate differential expression of CXCL5 in monocytes and osteoclasts and suggest a role for this chemokine in osteoclast biology. However, additional studies that are beyond the scope of this thesis work would be required to delineate a role for CXCL5 in osteoclast biology.

174 Chapter 7

General discussion

175 General Discussion

7.1 Overview

The work presented in this thesis evaluated the differential gene expression of a discrete selection of genes in divergent myeloid cell types. The common progenitor of two distinct and specialised myeloid cell types allowed the in vitro differentiation of these cells in such a way that changes in gene expression, as a result of polarised cell differentiation, could be evaluated alongside the characterisation of these cells with respect to their known phenotypic and functional attributes. The peripheral blood monocyte was chosen as a precursor for differentiation of myeloid dendritic cells and osteoclasts in vitro. These differentiation systems have been extensively documented in the literature and as such, cell populations were derived that are representative of their in vivo counterparts and suitable for further study. Firstly, immature dendritic cell populations were obtained by culturing peripheral blood monocytes with GM-CSF and IL-4; these populations expressed low levels of the co-stimulatory molecule CD80, were negative for the co-stimulatory molecule CD86, expressed medium levels of the class II MHC molecule HLA-DR and were negative for the dendritic cell maturation marker CD83. Mature dendritic cells were isolated by further differentiating the immature cells with monocyte-conditioned medium, resulting in cells with high levels of CD80, CD86, HLA-DR and CD83. Consistent with their roles in vivo, immature

DCs demonstrated high antigen uptake capacity and moderate ability to stimulate allogeneic T cells, whereas mature DCs had decreased antigen uptake function and increased ability to stimulate proliferation and IFN-y production by allogeneic T cells. Osteoclast populations were generated by culturing peripheral blood monocytes in the presence of M-CSF and RANKL; these cell populations contained a high percentage of TRAP-positive / aV(33-positive multinucleated cells with detectable levels of mRNA encoding RANK (the receptor for the osteoclast differentiation factor RANKL) and calcitonin receptor (a receptor routinely used in the identification of

176 osteoclasts). Functionally, the differentiated osteoclasts were capable of bone resorption upon reseeding onto dentine slices.

The differential expression of chemokines, chemokine receptors, related GPCRs and a number of other GPCR ligands were evaluated using these populations of monocytes, osteoclasts, immature

and mature DCs with a custom synthesised gene microarray. The gene array and protein analysis of chemokine production by DCs showed good correlation, where mRNA levels were present, protein was either constitutively produced or could be induced by further activation of the DCs. This study demonstrated constitutive production of the homeostatic chemokines CCL18 (PARC) in immature

DCs and CCL17 (TARC) and CCL22 (MDC) in mature DCs, consistent with the role of CCL18 in attracting naïve T cells (Adema et al., 1997) and CCL17 and CCL22 in the formation of lymphocyte-dendritic cell clusters at sites of inflammation (Katou et al., 2001). The inflammatory

chemokines CCL13 (MCP-4) and CCL23 (CK-^8) were demonstrated in immature DCs without the requirement of further stimulus; CCL13 is widely expressed at disease sites and involved in recruitment of inflammatory cells (Miotto et al., 2001). CCL23 binds to the CCRl receptor and is a potent activator of monocytes (Nardelli et al., 1999b). Other inflammatory chemokines, such as

CCL3 (MIP-la), CCL4 (MIP-ip) and CCL15 (MIP-1Ô) were induced in immature DCs with

inflammatory mediators. This is in accordance with the role of these inflammatory chemokines as

inducible chemokines, functioning through the receptors CCRl, CCR3 and CCR5 and recruiting

macrophages, lymphocytes, eosinophils, dendritic cells and NK cells to sites of infection (Driscoll,

1994; Cook, 1996). The fact that the expression profiles for chemokines and chemokine receptors in

DCs are consistent with published reports validates not only the significance of transcript

expression, as determined by the custom gene microarray, but further validates the in v/?ro-derived

dendritic cell populations used in this study. The expression of chemokines and chemokine

receptors by osteoclasts was mostly unremarkable, but they showed a decrease in several monocyte-

derived chemokines. Expression of CCL3, CCL4 and CCRl in these human in v//ro-derived

177 osteoclasts requires further confirmation. However, differential expression of one chemokine gene,

CXCL5, was detected in osteoclasts and a detailed investigation is detailed in Chapter 6 of this thesis.

The gene microarray analysis also revealed differential mRNA expression of the recently described

G protein-coupled receptor GPR105 (Chambers et al., 2000) in immature DCs. Expression of

GPR105 and functional activation of this receptor with its putative agonist ligand, UDP-glucose, as well as the consequences of activating this receptor on immature DCs were investigated. The expression pattern of GPR105 in monocytes, immature and mature DCs was confirmed by Northern blot analysis. UDP-glucose was corroborated as an agonist ligand for this receptor by the demonstration of a calcium flux and EC50 analysis in a HEK293 cell line co-transfected with

GPR105 and the promiscuous G protein subunit G al 6. Investigation of calcium responses to UDP- glucose in either monocytes or in viYro-derived immature and mature DC confirmed expression of a functional UDP-glucose receptor on immature DCs only. Treatment of immature DCs with UDP- glucose revealed a gene expression profile, as represented by screening with the custom gene microarray, synonymous with DC maturation. These DCs also displayed the phenotype of mature

DCs, with respect to protein expression of the co-stimulatory molecules CD80 and CD86, CD83 and HLA-DR and demonstrated a down-regulation in endocytic ability, again paralleling DC maturation. However, analysis of the UDP-glucose preparations used in these initial experiments revealed endotoxin contamination and further experimentation using these contaminated UDP- glucose preparations revealed a correlation between up-regulation of the co-stimulatory molecule

CD86 and the level of endotoxin contamination, with a subsequent neutralisation of these effects in the presence of the endotoxin-binding agent polymyxin B. The endotoxin contamination had no effect on the UDP-glucose-induced calcium flux detected in immature DC, as demonstrated by an equivalent calcium response in the presence of polymyxin B. Further experiments using endotoxin- free UDP-glucose preparations from ICN Biomedicals revealed an increased CD86 expression in

178 immature DCs in some donors, with a correlation between this UDP-glucose-induced CD86 up- regulation and the magnitude of a UDP-glucose-induced calcium response. Experiments were also performed to test if endotoxin-free UDP-glucose could enhance EPS-induced effects in DCs. EPS- activated DCs exhibited enhanced CD86 and CD83 expression with a decrease in endocytic ability, although these effects were not as potent as those seen with MCM. The addition of UDP-glucose further enhanced CD86 expression in the presence of EPS, whereas no effect was evident on CD83 expression and endocytic ability. Immature DCs treated with endotoxin-free UDP-glucose demonstrated enhanced allogeneic T cell activation potential with induction of IFN-y production, both in the presence and absence of EPS. While an unequivocal role for GPR105 in DC maturation was not determined by our studies, GPR105 does appear to be specifically expressed by immature

DCs and to be physiologically active on these cells, as evidenced by the calcium flux data. To further substantiate that UDP-glucose-mediated calcium fluxes are functioning through GPR105, the pertussis toxin sensitivity of the G proteins coupled to this receptor was exploited (Chambers et al., 2000). Pertussis toxin pre-treatment of immature DCs resulted in a decrease of the UDP- glucose-induced calcium flux; however the signal was not completely abolished. This demonstrates that the calcium signal obtained in response to UDP-glucose is, at least in part, working through pertussis toxin sensitive G proteins. As previously discussed, it is possible for G protein-coupled receptors to associate with different G protein classes and therefore display incomplete sensitivity to pertussis toxin (Haribabu et al., 1999; Ahamed and Ali, 2002). Moreover, HEK293 cells co­ expressing GPR105 and Gaie demonstrated a calcium flux in response to necrotic cell supernatants.

As discussed previously, the release of UDP-glucose on mechanical stimulation of cells is low but levels remain stable for several hours. This inefficient metabolism of UDP-glucose may allow for critical thresholds of UDP-glucose to be ahieved under certain circumstances. Perhaps GPR105 signalling requires this critical threshold of ligand, thereby avoiding inappropriate activation of

DCs. In conclusion, these data demonstrate a correlation between UDP-glucose-induced signalling

179 in immature DCs and the expression profile of GPR105. A possible role for UDP-glucose in DC maturation is described, although further investigation is required to ascribe a functional response to this receptor in DCs.

The expression of CXCL5 in osteoclast cultures was evaluated alongside CXCL8 (IL-8), a CXC

chemokine previously described in osteoclasts (Rothe et al., 1998). Firstly, high levels of CXCL5 protein production were demonstrated to correlate with the formation of multinucleated cells in in v//ro-derived osteoclast differentiation cultures; other CXC chemokines, CXCLl, CXCL7 and

CXCL8 were found at consistent levels throughout the time course. Immunocytochemistry

confirmed multinucleated osteoclasts as a source of both CXCL5 and CXCL8 protein production in these cultures. Chemotaxis experiments demonstrated powerful chemoattraction of neutrophils to

osteoclast supernatants and this activity could be blocked with an antibody to CXCL8. Although an

antibody to CXCL5 could only significantly block neutrophil chemotaxis to one of three osteoclast

culture supernatants, chemotaxis was reduced in all culture supernatants and the poor contribution

of CXCL5 to neutrophil migration may make significance difficult to attain. Chemotaxis

experiments utilising endothelial cells, where CXCL5 is equipotent to CXCL8 (Strieter et al.,

1995), were not possible due to the large serum component of the osteoclast supernatants. Lastly,

the effects of angiogenic and inflammatory stimuli on the production of CXCL5 in osteoclast

cultures were investigated. Inflammatory stimuli increased CXCL5 production in accordance with

an increase in multinucleated cell formation in these cultures; in contrast the angiogenic stimulus

VEGF had no effect. The data reported demonstrates differential expression of CXCL5 in

osteoclasts versus monocytes and suggests a role for this chemokine in osteoclast biology.

7.2 Discussion

Despite the reported successful differentiation of DCs from myeloid and lymphoid precursors in

vitro, the differentiation pathways that generate DCs in vivo remain largely unknown (Ardavin et

180 al., 2001). However recent experiments have been able to track the differentiation of DCs in vivo, through the adoptive transfer of green fluorescent protein-labelled monocytes from two distinct monocyte subsets, which has added to the weight of evidence that at least some in vivo DC populations are monocyte-derived (Geissmann et al., 2003). Although these experiments directly demonstrate in vivo differentiation of DCs from monocyte precursors in an experimental mouse, the authors were able to identify putative human counterparts of these distinct monocyte subsets. Some of the differences between results in human and mouse DCs may result from differences in the isolation or in vitro differentiation techniques used. With the knowledge of the existence of tissue- specific DC populations, as well as distinct stages of maturation/activation of DCs, it is becoming clear that human and mouse DC subpopulations share many phenotypic and functional characteristics (Shortman and Liu, 2002; O'Keeffe et al., 2003). Although no direct in vivo evidence exists for the differentiation of osteoclasts from monocytes, perhaps the most compelling data comes from the study of the op/op mice, which have a mutation in the gene encoding the critical macrophage growth and survival factor M-CSF, also called colony-stimulating factor type 1 (CSF-

1) (Cecchini et al., 1994). These mice are deficient in several mononuclear phagocyte subpopulations, including osteoclasts. However, it remains a pertinent point that the most appropriate cells for the gene transcript expression studies described in this thesis are fully differentiated cells purified from blood or tissues. However, large numbers of purified cells from human blood or tissue may be difficult to obtain, restricting studies to a more limited number of genes, such as those identified in this study.

Microarray technology has already provided an unanticipated and novel finding with respect to DC biology. Granucci et al. (Granucci et al., 2001) performed gene expression analysis on immature murine DCs stimulated at different time-points with Gram-negative bacteria. In their study a gene microarray representing 11,000 genes and ESTs was screened. Gene clustering methods revealed differential expression of approximately 3,000 genes and ESTs, including TNF-a, CCL3, CCL4

181 and IL-12, which are well-established markers for DC activation. Unexpectedly, expression of the

IL-2 gene in DCs following bacterial challenge was also observed. IL-2 has long been recognised as a key molecule in the growth and proliferation of activated T cells. The impaired ability of DCs infected with cytomegalovirus to produce IL-2 and subsequently to prime alloreactive T cells has since been demonstrated (Granucci et al., 2002). This thesis describes the use of a gene microarray to analyse differential expression of mRNA transcripts encoding a discrete selection of genes in human monocytes, monocyte-derived immature and mature DCs, as well as monocyte-derived osteoclasts. There are no reports in the literature of gene microarray analyses involving in vitro- derived osteoclasts. However, other studies comparing the transcript profiles of human monocyte- derived immature and mature DCs have been reported; Dietz et al. utilised a DNA microarray containing fragments for 4,110 known genes, demonstrating differential transcript expression in 369 genes (Dietz et al., 2000). The most impressive data set from their study was the upregulation of mRNA transcripts encoding the chemokine and chemokine receptor genes CCL17, CCL22 and

CCR7 upon DC maturation. These findings are reinforced by the results obtained in this thesis.

Other DNA microarray data presented by Dietz et al. allowed the authors to propose novel hypotheses for maturation-associated changes in dendritic cells and their interactions with the microenvironment, such as the proposal that the upregulated expression of membrane bound TGF-a upon DC maturation could aid in the attachment of DCs to stromal elements in the lymph nodes, or that TGF-a released from DCs might stimulate stromal cells over a longer range. Le Naour et al.

(Le Naour et al., 2001) compared mRNA transcript profiles of immature and mature DCs and their monocyte precursors, by hybridising cDNA samples to DNA microarrays comprising oligonucleotides specific for 6,300 known genes. A total of 255 genes, mostly not previously associated with DCs, were found to be differentially regulated during DC differentiation and maturation, including genes related to cell adhesion, motility and regulation of the immune response. The authors combined the gene expression profiling with a proteomics approach, whereby

182 protein analysis utilising two-dimensional gel electrophoresis and mass spectrometry identified differentially expressed proteins. This provided additional information that could not be obtained at the mRNA level, due to a poor correlation between mRNA expression and protein synthesis, or post-transcriptional modifications that may result in several spliced isoforms generated from precursor RNA. Another consideration when analysing mRNA transcript expression data revealed by microarray profiling is that at any particular time frame highly regulated transcripts may not be detected. As mentioned in the discussion of Chapter 4, transcripts encoding the inflammatory chemokines CCL3 and CCL4, although widely reported to be produced by mature / activated DCs

(Sallusto et al., 1999), were not seen in mature DCs. It is also important to recognise that degradation of mRNA is tightly regulated and many mammalian inducible mRNA species have been shown to have a short half-life due to the presence of regulatory elements in their sequences that target them for efficient degradation (Bevilacqua et al., 2003). Conversely, the presence of an mRNA species does not ensure synthesis of the corresponding protein, as translation is not only regulated by universal mechanisms but more specific mechanisms exist that influence individual mRNAs (Cazzola and Skoda, 2000). It remains highly pertinent that all transcript expression data generated by gene microarray profiling should be confirmed, by a second or third technique, preferably by identification of the expressed protein.

Critics of microarray profiling, at least where thousands of genes are screened at once, have referred to these data driven approaches as fishing expeditions. The approach taken in this thesis was to analyse a discrete selection of genes in related cell types. The main hypothesis being that by using microarray technology to examine the distinct gene expression patterns of highly related, yet fimctionally distinct cell types, one may identify genes with functional relevance to the specific cell types. Candidate genes for further work included the orphan GPCR, H963, which showed upregulation of message on the mature DC population. Little is known about H963 except that its closest human ortholog is the platelet activating factor receptor (Jacobs et al., 1997). This

183 unconfirmed finding of expression in mature DCs might suggest a role in antigen presentation or T cell activation by these cells. In fact, platelet-activating factor receptor is present on immune cells and its ligand, platelet activating factor, is secreted by immune cells (Camussi et al., 1990). Thus, it is possible to postulate an immunological role for the expression of the ortholog H963 on mature dendritic cells. Another candidate for further work was the CCL15 gene, mRNA encoding CCL15 was upregulated in immature DCs. Production of this chemokine on stimulation of immature dendritic cells with IFN-y and LPS or SAC was confirmed in Chapter 4 of this thesis. The characterisation of CCL15 as a chemoattractant for monocytes, eosinophils, T cells, dendritic cells, and neutrophils, mediated through CCRl and CCR3, is consistent with the production of this chemokine by immature DCs at sites of inflammation (Nardelli et al., 1999b). Two further genes, previously uncharacterised in the cell type expressed, were extensively evaluated.

The transcript encoding GPR105 was demonstrated in immature DCs by DNA microarray profiling and confirmed by Northern blot analysis. Specific antibodies to GPR105 were not available during this study, preventing direct confirmation of the expressed GPR105 protein. Lee et al. (Lee et al.,

2003) have recently described the use of a polyclonal anti-GPRl 05-peptide antibody to demonstrate the expression of GPR105 on a subset of haemopoietic cells within the foetal bone marrow, by immunofiuoresence and flow cytometry. Confirmation of GPR105 expression on immature DCs using this antibody would support the work reported in this thesis. However, demonstration of

UDP-glucose-initiated signalling in immature DCs and not in monocytes and mature DCs, in which

GPR105 transcript expression is decreased, adds support for the functional expression of GPR105 on immature DCs. A neutralising antibody to GPR105 could also be used to confirm that UDP- glucose initiated signalling is mediated via GPR105 and not via another, as yet unidentified, UDP- glucose receptor. The emergence of UDP-glucose as an extracellular signalling molecule

(Lazarowski et al., 2003) and evidence that DC maturation can be induced through exposure to necrotic tumour cells (Sauter et al., 2000), lead to the proposal that UDP-glucose could be acting as

184 a danger signal. The data presented in Chapter 5 of this thesis demonstrating activation of GPR105

by necrotic cell supernatants adds to the possibility that UDP-glucose, in concert with other signals,

may be promoting the activation / maturation of DCs on its release. Confirmation of an increase in

the concentration of UDP-glucose in the necrotic cell supernatants would strengthen this argument

and assays to measure this have recently been developed (Lazarowski et al., 2003). The enhanced

levels of UDP-glucose and other glycolytic metabolites in tumour cells (Eigenbrodt et al., 1992)

suggest a role for GPR105 expression on immature DCs in tumour surveillance. Another possibility, that GPR105 could be involved in dendritic cell homing, warrants investigation. Indeed

GPR105 has some homology to chemokine receptors (20 - 30% amino acid homology) and has the

sequence motif DRYYKIV, located at the end of transmembrane III (Lee et al., 2003), similar to the

DRYLAIV motif, a signature sequence for chemokine receptors (Youn et al., 1997a). The enhanced metabolism seen in cancer tissues (Eigenbrodt et al., 1992) is also likely to be seen in other actively proliferating cell systems, such as embryonic tissues (Ito et al., 1999). Perhaps lymphoid tissues such as lymph node, spleen or thymus, to which immature DCs home, contain high levels of UDP- glucose as a result of the actively proliferating cells present during times of immune activation or development. In fact, Lee et al. (Lee et al., 2003) used conditioned media from bone marrow stroma to attract GPR105 expressing cells, and the source of these stromal cells, the bone marrow, can also be described as an actively proliferating cell system.

The second gene that warranted further investigation was CXCL5, demonstrated in osteoclasts by gene microarray profiling and confirmed by showing presence of secreted CXCL5 protein by both

ELISA and immunocytochemistry. The presence of CXCL5 in primary in v/vo-derived osteoclasts awaits demonstration. However, techniques for the isolation of in v/vo-derived osteoclasts commonly use cells from chickens or rabbits, where orthologs for CXCL5 have not been described.

Demonstration of human osteoclasts expressing CXCL5 by immunohistochemistry remains a possibility. Chemotaxis experiments were utilised to exhibit the functional activity of osteoclast-

185 produced CXCL5 and the possibility that this chemokine is involved in the process of angiogenesis was discussed. Demonstration of CXCL5-producing osteoclasts in close proximity with endothelial cells during bone remodelling, would suggest a role for osteoclasts in attracting these cells to an area of bone where new blood vessel formation is required. The effect of VEGF on the production of CXCL5 by osteoclasts was examined in Chapter 6 of this thesis; no increase in CXCL5 production was seen. However, recent reports demonstrate the upregulation of RANK, the RANKL receptor, on endothelial cells in response to VEGF (Min et al., 2003) and the enhancement of endothelial cell capillary tube formation in response to RANKL (Kim et al., 2002). It would therefore be interesting to evaluate the effects of VEGF with respect to RANK expression on osteoclasts or osteoclast precursors. Although no direct relationship could be demonstrated between

CXCL5 production from osteoclasts and the angiogenic factor VEGF in this non-exhaustive study, it is clear that there are many relationships emerging between osteoclastogenesis and angiogenesis.

Certainly, unequivocal evidence of CXCL5 induced endothelial cell migration to the site of bone turnover would be an attractive addition to this story. In addition, the inflammatory cytokines TNF- a and IE-1(3 were shown to increase CXCL5 production from osteoclast cultures in Chapter 6 of this thesis, albeit maybe just as a result of increased osteoclastogenesis. Equally CXCL5 has recently been shown to induce TNF-a and IE-1(3 expression by endothelial cells (Chandrasekar et al., 2003). TNF-a and IL-ip are capable of promoting further osteoclastogenesis, suggesting a self­ regulating role for osteoclast-produced CXCL5. Lastly, RANKL production from endothelial cells in response to TNF-a has also been reported (Collin-Osdoby et al., 2001). Taken together, this information supports the concept of a network of interactions between osteoclasts and endothelial cells in bone formation and angiogenesis. Furthermore a role for CXCL5 in angiogenesis in inflamed tissue in rheumatoid arthritis has been proposed (Koch et al., 2001). The data generated in this thesis identifies osteoclasts as possible contributors to the presence of CXCL5 in these diseased tissues and further implicates an important role for CXCL5 in bone remodelling and inflammation.

186 7.3 Summary

Microarray gene technology applied to specialised in v/fro-derived myeloid cell types has lead to the investigation of two distinct genes in divergent areas of biology. Perhaps the greatest point for discussion is that cells arising from a single peripheral blood precursor are involved in such diverse functions as the initiation of an immune response and bone remodelling. At the outset of this thesis the hypothesis was to use microarray technology to examine the distinct gene expression patterns of highly related, yet functionally distinct cell types, in an effort to identify genes with functional relevance to the specific cell type. The characterisation of functionally active GPR105 on immature

DCs and the demonstration of biologically active CXCL5 production by osteoclasts validate this hypothesis.

One further aim for this thesis was to demonstrate a role for a crucial set of proteins, chemokines and chemokine receptors, in osteoclast biology and compare this to the well-documented role for this family in highly migratory dendritic cells. Perhaps the in situ nature of osteoclast formation should have highlighted the probability of a less important role for this family in osteoclasts. What emerged from the DNA microarray data was the novel and differential expression of two genes;

GPR105 in immature dendritic cells and CXCL5 in osteoclasts. As neither of these findings were expected, this demonstrates the power of microarray profiling, where numerous, indeed even thousands of genes may be investigated at one time, resulting in novel findings. The expression of

GPR105 on immature DCs has led to the investigation of a putative role for this receptor in the initiation of an immune response, with confirmation of GPR105 mRNA transcript expression, as well as demonstration of signalling via a G protein-coupled receptor in response to the GPR105 agonist UDP-glucose. The expression of mRNA transcripts encoding CXCL5 in osteoclasts has resulted in a proposed role for osteoclast-secreted CXCL5 in angiogenesis, with confirmation of

CXCL5 protein production by in v//ra-derived osteoclasts.

187 Appendix 1

Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 day 0 day 9

FPRL1 -9 + / - 4 -1 + / - 2 -4 + / - 2 -2 + / - 2 -2 + / - 1 P2Y5 15 + / - 2 15 + / - 2 3 + / - 1 8 + / - 1 6 + / - 2 CML2 24 + / - 6 16 + / - 6 7 + / - 3 14 + / - 4 17 + / - 3 0GR1 short 1 + / - 3 3 + / - 2 -1 + / - 1 0 + / - 1 1 + / - 1 0GR1 long 13 + / - 3 5 + / - 1 6 + / - 2 8 + / - 2 9 + / - 2 GPR18 7 + / - 1 3 + / - 1 8 + / - 2 5 + / - 1 3 + / - 1 H963 short -8 + / - 1 -7 + / - 3 -4 + / - 2 -7 + / - 1 -6 + / - 1 H963 long 4 + / - 2 10 + / - 2 43 + / - 7 1 + / - 1 3 + / - 1 TDAG8 9 + / - 4 8 + / - 1 2 + / - 2 31 + / - 8 11 + / - 2 C3AR 19 + / - 5 26 + / - 6 -2 + / - 2 25 + / - 3 34 + / - 7 P2Y9 -11 + / - 3 -6 + / - 1 -2 + / - 1 0 + / - 1 -1 + / - 1 EBI2 15 + / - 7 17 + / - 2 15 + / - 3 22 + / - 8 11 + / - 3 HSGPCP 8 + / - 4 4 + / - 3 9 + / - 3 9 + / - 2 11 + / - 2 GPR105 174 + / - 50 255 + / - 27 43 + / - 16 18 + / - 2 14 + / - 1 GPR34 short -5 + / - 5 3 + / - 2 3 + / - 3 1 + / - 1 4 + / - 1 GPR34 long -10 + / - 6 -9 + / - 3 -4 + / - 2 -3 + / - 1 6 + / - 4

GPR38 -3 + / - 1 0 + / - 1 2 + / - 1 3 + / - 1 2 + / - 1 GPR15 4 + / - 1 5 + / - 1 1 + / - 0 1 + / - 1 7 + / - 1 GPR31 1 + / - 3 1 + / - 0 1 + / - 1 1 + / - 1 1 + / - 1 GPR32 short 12 + / - 1 17 + / - 4 11 + / - 3 32 + / - 2 32 + / - 2 GPR32 long 6 + / - 3 9 + / - 3 10 + / - 2 28 + / - 3 24 + / - 2 CRTH2 -1 + / - 4 2 + / - 2 0 + / - 1 3 + / - 1 4 + / - 2 GPR44 2 + / - 12 -3 + / - 1 1 + / - 3 1 + / - 1 1 + / - 1 GPR16 56 + / - 7 41 + / - 12 27 + / - 6 108 + / - 10 38 + / - 4 GPR115 0 + / - 1 6 + / - 2 9 + / - 1 3 + / - 2 3 + / - 1 GPR128 23 + / - 10 14 + / - 3 19 + / - 3 33 + / - 10 7 + / - 2 GPR132 4 + / - 2 10 + / - 1 5 + / - 0 19 + / - 1 .9 + / - 2 CNR2 7 + / - 3 5 + / - 1 6 + / - 1 8 + / - 1 7 + / - 1 PTAFR 73 + / - 5 63 + / - 8 25 + / - 4 83 + / - 5 65 + - 7

Orphan G-protein 7TM receptors, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

188 Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 day 0 day 9 Thrombin 12 + - 1 10 + / - 3 2 + - 1 4 + / - 1 14 + / - 6 Trypsin 1 8 + - 5 8 + / - 2 5 + - 1 8 + / - 6 5 + / - 1 Trypsin 2 26 + - 6 22 + / - 4 15 + - 2 15 + / - 2 15 + / - 3 Trypsin 3 10 + - 3 16 + / - 3 14 + - 6 8 + / - 3 4 + / - 4 Trypsin 4 16 + - 5 16 + / - 3 13 + - 2 9 + / - 2 26 + / - 5

Tryptase a 22 + - 6 22 + / - 6 20 + - 3 17 + / - 3 44 + / - 7

Tryptase b 34 + - 3 33 + / - 7 20 + - 4 18 + / - 4 46 + / - 7 Elastase 2A 22 + - 6 16 + / - 6 9 + - 2 19 + / - 4 20 + / - 3 Elast 28 Short -9 + - 2 -2 + / - 2 6 + - 6 -2 + / - 3 -4 + / - 2 Elast 28 Long 8 + - 4 7 + / - 3 6 + - 2 11 + / - 2 14 + / - 3 Elastase 3A 21 + - 6 18 + / - 6 11 + - 3 18 + / - 3 12 + / - 3 Elastase 38 27 + - 7 19 + / - 6 16 + - 3 21 + / - 3 22 + / - 5 Caldecrin 23 + - 6 17 + / - 5 14 + - 3 12 + / - 2 15 + / - 1 Neuromedin_U 1 + - 6 0 + / - 0 1 + - 1 0 + / - 1 0 + / - 1

P2Y2 4 + 2 6 + / 3 31 + - 6 20 + / - 6 3 + / - 1 PAR1 66 + - 18 77 + / - 15 30 + - 3 78 + / - 4 81 + / - 5 PAR2 -1 + - 3 0 + / - 1 -3 + - 1 4 + / - 3 0 + / - 0 PAR3 39 + - 2 37 + / - 3 31 + - 2 69 + / - 6 62 + / - 5 PAR4 39 + - 7 24 + / - 11 25 + - 9 34 + / - 5 17 + / - 4 C5aR 30 + - 10 37 + / - 6 6 + - 2 31 + / - 3 37 + / - 9 C3aR 30 + - 8 34 + / - 5 5 + - 2 21 + / - 1 34 + / - 8 Edg1 3 + - 2 6 + / - 2 7 + - 2 4 + / - 3 7 + / - 3 Edg2 -10 + - 2 -5 + / - 2 -2 + - 2 -4 + / - 2 -2 + / - 1 Edg3 10 + - 4 6 + / - 2 26 + - 4 8 + / - 1 8 + / - 1 Edg4 11 + - 2 10 + / - 1 3 + - 1 19 + / - 3 4 + / - 2 Edg5 68 + - 6 46 + / - 8 64 + - 9 71 + / - 4 43 + / - 5 Edg6 25 + - 6 20 + / - 3 15 + - 2 43 + / - 5 8 + / - 1 Edg7 25 + - 13 10 + / - 3 6 + - 1 8 + / - 3 10 + / - 2 EdgS 4 + - 6 8 + / - 2 6 + - 1 7 + / - 1 9 + / - 2 082 2 + - 2 5 + / - 2 4 + - 2 6 + / - 2 5 + / - 1 EMR1 3 + - 1 3 + / - 1 2 + - 1 22 + / - 3 4 + / - 1 CD97 33 + - 2 20 + / - 4 9 + - 1 93 + / - 6 41 + / - 3

Known GPCR and GPCR ligands of interest, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

189 Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 day 0 day 9 CCL1 short -27 + / - 19 -22 + / - 7 -15 + / - 3 1 + / - 1 0 + / - 1 CCL1 long 0 + / - 2 1 + / - 1 -4 + / - 2 -5 + / - 2 4 + / - 1 CCL2 34 + / - 15 33 + / - 9 326 + / - 38 -7 + / - 4 11 + / - 2 CCL3 34 + / - 10 40 + / - 10 5 + / - 2 261 + / - 60 48 + / - 10 SCYA3L 111 + / - 45 166 + / - 46 13 + / - 10 268 + / - 60 56 + / - 8 CCL4 33 + / - 5 29 + / - 6 -8 + / - 4 189 + / - 57 19 + / - 4 CCL5 Short 13 + / - 4 22 + / - 4 29 + / - 7 34 + / - 2 44 + / - 2 CCL5 Long 78 + / - 4 59 + / - 10 53 + / - 13 157 + / - 13 143 + / - 11 CCL7 Short -4 + / - 1 -2 + / - 1 -1 + / - 1 -3 + / - 2 6 + / - 5 CCL8 -5 + / - 3 -1 + / - 2 -2 + / - 2 -2 + / - 1 9 + / - 2 CCL11 67 + / - 2 53 + / - 8 62 + / - 12 134 + / - 11 168 + / - 20 CCL13 192 + / - 40 251 + / - 26 33 + / - 17 3 + / - 1 4 + / - 1 CCL14 58 + / - 12 57 + / - 9 22 + / - 12 22 + / - 5 13 + / - 3 CCL15 97 + / - 29 74 + / - 22 15 + / - 5 -6 + / - 2 -3 + / - 2 CCL16 10 + / - 6 6 + / - 3 11 + / - 2 41 + / - 3 37 + / - 4 CCL17Short 77 + / - 3 129 + / - 29 488 + / - 20 7 + / - 1 4 + / - 1 CCL17 Long 140 + / - 6 235 + / - 29 749 + / - 33 4 + / - 2 8 + / - 1 CCL18 304 + / - 134 519 + / - 120 58 + / - 39 15 + / - 6 21 + / - 6 CCL19Short -2 + / - 3 -1 + / - 1 -6 + / - 3 0 + / - 1 4 + / - 1 CCL19 Long 10 + / - 9 -4 + / - 5 38 + / - 8 16 + / - 3 21 + / - 3 CCL20 3 + / - 3 -1 + / - 1 -14 + / - 4 35 + / - 14 2 + / - 1 CCL21 44 + / - 13 29 + / - 9 15 + / - 3 24 + / - 3 20 + / - 3 CCL22 49 + / - 16 47 + / - 5 343 + / - 37 11 + / - 2 13 + / - 3 CCL23 149 + / - 28 111 + / - 27 21 + / - 10 1 + / - 2 2 + / - 2 CCL24 55 + / - 2 91 + / - 21 68 + / - 7 33 + / - 3 20 + / - 4 CCL25 12 + / - 5 9 + / - 4 6 + / - 2 9 + / - 3 17 + / - 2 CCL26 0 + / - 4 -2 + / - 1 1 + / - 1 1 + / - 3 4 + / - 1 CCL27 8 + / - 5 10 + / - 3 7 + / - 1 3 + / - 2 8 + / - 2 CCL28 59 + / - 4 60 + / - 4 48 + / - 3 88 + / - 3 44 + / - 6 XCL1 13 + / - 2 13 + / - 1 12 + / - 2 14 + / - 1 15 + / - 2 CX3CL1 19 + / - 8 17 + - 7 6 + / - 2 2 + / - 1 2 + / - 1

CC and CX3C chemokines, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

190 Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 day 0 day 9 CCR1 43 + / - 6 58 + / - 5 3 + / - 5 38 + / - 6 55 + / - 6 CCR2 9 + / - 4 2 + / - 5 0 + / 35 + / - 6 3 + / - 1 CCR3 Short -1 + / - 0 0 + / - 1 0 + / -1 + / - 1 0 + / - 1 CCR3 Long -3 + / - 1 2 + / - 4 0 + / -2 + / - 2 4 + / - 1 CCR4 -5 + / - 4 1 + / - 1 0 + / -2 + / - 3 3 + / - 1

CCR5 42 + / - 7 27 + / - 4 2 + / 4 + / - 2 24 + / - 3 CCR6 10 + / - 6 5 + / - 2 3 + / 1 + / - 1 4 + / - 1 CCR7 11 + / - 3 11 + / - 2 117 + / - 12 13 + / - 1 12 + / - 2 OCRS short 0 + / - 1 2 + / - 1 1 + / 5 + / - 1 3 + / - 2 CCR8 long 3 + / - 2 4 + / - 2 -2 + / -4 + / - 2 6 + / - 1 CCR9 0 + / - 2 1 + / - 1 0 + / 6 + / - 1 3 + / - 1

CCR10 50 + / - 7 41 + / - 13 36 + / - 10 54 + / - 3 37 + / - 5 CCR11 5 + / - 5 4 + / - 1 8 + / - 4 7 + / - 2 8 + / - 1 CXCR1 2 + / - 2 3 + / - 2 0 + / - 1 2 + / - 1 5 + / - 1 CXCR2 36 + / - 10 24 + / - 5 8 + / - 2 18 + / - 2 8 + / - 2

CXCR3 46 + / - 8 30 + / - 11 29 + / - 7 40 + / - 5 44 + / - 2

CXCR4 48 + / - 1 58 + / - 3 76 + / - 6 89 + / - 15 70 + / - 7 CXCR5 5 + / - 1 13 + / - 1 11 + / - 1 7 + / - 1 11 + / - 1 CX3CR1 -3 + / - 3 4 + / - 2 2 + / - 1 25 + / - 5 5 + / - 1

CCXCR1 12 + / - 7 17 + / - 5 11 + / - 1 19 + / - 3 18 + / - 4 D6 15 + / - 4 15 + / - 3 1 + / - 2 9 + / - 3 19 + / - 2 DARC 27 + / - 4 16 + / - 5 9 + / - 1 5 + / - 5 18 + / - 3 CCRL2 -1 + / - 2 6 + / - 2 2 + / - 1 6 + / - 3 9 + / - 2 hm74 13 + / - 8 7 + / - 3 4 + / - 2 25 + / - 8 5 + / - 1

RDC1 8 + / - 2 11 + / - 3 9 + / - 0 9 + / - 2 9 + / - 2 TYMSTR -5 + / - 2 -3 + / - 2 3 + / - 3 -4 + / - 3 -2 + / - 2

Chemokine Receptors, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

191 Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 day 0 day 9 CXCL1 -1 + / - 2 2 + / - 2 6 + / - 2 19 + / - 10 24 + / - 7 CXCL2 1 + / - 1 6 + / - 3 3 + / - 3 58 + / - 17 13 + / - 3 CXCL3 27 + / - 1 12 + / - 5 17 + / - 6 44 + / - 9 18 + / - 2 CXCL4 19 + / - 4 15 + / - 3 12 + / - 3 18 + / - 2 15 + / - 2 CXCL5 -2 + / - 3 2 + / - 4 6 + / - 4 -3 + / - 1 46 + / - 14 CXCL6 -2 + / - 2 2 + / 1 + / - 1 -11 + / - 4 8 + / - 3 CXCL7 -1 + / - 3 2 + / 4 + / - 2 38 + / - 9 77 + / - 30 CXCL8 0 + / - 2 1 + / 7 + / - 4 175 + / - 45 23 + / - 4 CXCL9 -10 + / - 6 -2 + / - 4 0 + / - 1 3 + / - 1 0 + / - 1 CXCL10 0 + / - 1 1 + / -1 + / - 1 2 + / - 1 1 + / - 2 CXCL11 2 + / - 4 1 + / 0 + / - 1 0 + / - 1 1 + / - 1 CXCL12 -9 + / - 7 -9 + / - 2 -19 + / - 4 1 + / - 1 4 + / - 1 CXCL13 -8 + / - 4 -7 + / -28 + / - 3 -1 + / - 1 1 + / - 1 CXCL14 0 + / - 2 3 + / 4 + / - 1 4 + / - 2 3 + / - 1

Hu_genomic_75_ng 197 + / - 13 173 + / - 23 130 + / - 17 172 + / - 15 241 + / _ 52 Hu_genomic_75_ng 187 + / - 16 153 + / - 27 114 + / - 23 171 + / - 17 234 + / - 42 Hu_genomic_75_ng 177 + / - 14 142 + / - 26 103 + / - 18 167 + / - 12 250 + / - 46 Hu_genomic_75_ng 204 + / - 16 156 + / - 28 116 + / - 22 203 + / - 12 251 + / - 48 Hu_genomic_75_ng 218 + / - 6 189 + / - 29 147 + / - 21 249 + / - 20 308 + / - 40 H u_genom ic_7 5_ng 209 + / - 7 179 + / - 33 139 + / - 22 237 + / - 20 292 + / - 40 Hu_genomic_75_ng 160 + / - 14 133 + / - 18 97 + / - 13 154 + / - 9 228 + / - 50 Hu_genomic_75_ng 239 + / - 23 173 + / - 40 131 + / - 28 221 + / - 15 248 + / - 48 H u_genom ic_7 5_ng 227 + / - 5 195 + / - 35 156 + / - 27 258 + / - 18 284 + / - 31 Hu_genomic_75_ng 243 + / - 21 204 + / - 42 146 + / - 26 256 + / - 13 269 + / - 40

Cot-1 _1000_ng/ul 115 + / - 6 96 + / - 9 99 + / - 15 124 + / - 7 117 + / 18 Cot-1 _50_ng/ul 2 + / - 5 10 + / - 2 4 + / - 2 5 + / - 1 16 + / - 3 Cot-1 500 ng/ul 48 + / - 4 56 + / - 3 33 + / - 2 55 + / - 3 74 + / - 14

CXC-Chemokines, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridized to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

192 Immature DC Immature DC Mature DC Monocyte Osteoclast day 6 day 9 day 9 dayO day 9 Human mRNA for platelet-typejDhosphofmctokinase 93 + -15 75 + -4 47 + -4 28 + -4 84 + /-5 Human lysosomal giycosyiasparaginase gene 13 + -3 17 + - 1 14 + -1 17 + -3 24 + /-3 Human ETS2 oncogene 4 + - 2 9 + - 2 10 + - 2 54 + - 6 20 + /-4 Human capping protein a-mRNA partial cds 37 + -5 40 + -1 0 17 + -7 21 + -7 32 +/-6 Human mRNA for mitochon 60 + -5 60 + -4 38 + -3 50 + -3 56 + /-5 Human cytoplasmic l>actin gene 557 + - 12 514 + -31 670 + -33 657 + -36 475 +/-31 Unknown est hits only 2 + -5 6 + -3 6 + -2 6 + - 1 4 + /- 1 Human mitochondrial ADP/ADT translocator mRNA 23 + -7 30 + -7 17 + -4 22 + -8 46 + / - 6 Human superoxide dismutase mRNA 65 + -7 79 + -3 49 + -4 70 + -5 57 +/-8 Homo sapiens protein tyrosine kinase (Syk) mRNA 65 + -3 74 + -4 40 + -3 71 + -5 60 +/-6 Lipoamide dehydrogenase (NAD) 22 + -2 30 + -5 29 + -4 23 + -2 30 +1-2 Unknown -20 est only -6 + -5 3 + -2 4 + -2 6 + -2 8 +/-1 Human mRNA for lactate dehydrogenase-A 125 + -6 112 + -5 140 + -7 76 + -2 212 + / - 13 Human glutamate receptor 2 mRNA -4 + -3 0 + -4 -6 + -3 -9 + -2 1 + /-3 Human a-2-macroglotxjlin mRNA 77 + -21 125 + -22 40 + -7 22 + -5 59 + / - 14 Human glyceraldehyde 3-phosphate dehydrogenase 247 + -6 257 + -7 233 + -7 340 + - 13 622 +Z-36 DAD-1 pBSAS 43 + -4 42 + -5 31 + -2 40 + -5 44 + /- 4 SLUG-1 pBSAS -14 + -4 5 + -3 -1 + -5 -13 + -2 4 +1-2 Actin 5' fragment 160 + -25 165 + -26 188 + -30 121 + -27 135 +/-21 Actin Mddle fragment 197 + -24 181 + -22 217 + -26 185 + -21 200 +Z-28 Actin 3' Fragment 572 + -25 543 + -34 696 + -25 652 + -40 426 +Z-29 deo^ypusine synthase mRNA 44 + -2 43 + -6 45 + -3 47 + -4 58 +Z-3 eukaryotic initiation factor 2B-e 16 + -4 21 + -2 16 + - 1 20 + -3 15 +Z-4 90-kDa heat-shock protein gene 99 + -3 102 + -5 87 + -6 77 + -2 85 +Z-5 histone H2B.1 mRNA3'_end 31 + -5 39 + -4 47 + -8 41 + -2 85 +Z -8 ADP-ritosylation factor 1 gene 105 + -10 123 + -6 93 + -7 120 + -9 103 +1-7 enylyl cyclase associated protein 136 + -4 140 + -7 131 + -4 150 + -5 124 +Z-5 cytosolic malate dehydrogenase 93 + -3 115 + - 14 73 + -8 33 + - 1 77 +Z-8 haperonin protein (Tcp20) 46 + -2 50 + -4 45 + -5 56 + -4 41 +Z-5 glutamate dehydrogenase (GDH) 51 + -7 59 + -3 38 + -5 55 + -2 42 +Z-4 DNA repair helicase (ERCC3) 42 + -3 44 + -2 36 + -2 34 + -2 33 +1-2 endothelin-1 gene 13 + - 1 14 + - 1 9 + -2 11 + -2 13 + Z -1 (cytosin-5) methyltransferase 94 + -7 86 + -6 31 + -2 42 + -5 47 +Z-6 uroporphyrinogen 111 synthase mRNA 52 + - 1 54 + -2 48 + - 1 46 + -3 59 +1-2 calmodulin 183 + -8 192 + - 14 331 + -21 180 + - 15 111 + Z -19 MRL3 = mammalian ritxosome L3 26 + - 1 30 + -4 33 + -5 19 + -2 19+Z-3 ADP/ATP translocase mRNA 3' end done 176 + -10 160 + - 11 146 + -6 237 + - 10 198 +1-1 cytochrome tx>1 complex core protein 11 63 + -6 64 + -7 47 + -4 70 + -12 69 +Z -6 hnRNP C2 protein mRNA 108 + -5 109 + -4 96 + - 6 111 + -5 98 +Z-7

Housekeeping genes, array data summary table Monocytes were purified from PBMC. Dendritic cells were differentiated from monocytes by incubation with GM-CSF and IL-4. Day 6 immature DCs were incubated with or without MCM (mature DCs and immature DCs respectively) for 72 hours. Osteoclasts were derived from monocytes in the presence of M-CSF and RANKL for 9 days. RNA was extracted and reverse transcribed into ^^P-labelled cDNA, which was hybridised to a gene microarray. Scanned phosphorimage data was extracted using IMAGENE software and normalised to the average expression level of the housekeeping gene panel, with final expression levels expressed as % average of this panel. Data are means of 3 (day 6 immature DCs), 6 (monocyte, day 9 immature DCs and day 9 mature DCs) or 7 (osteoclast) experiments, SEMs are shown.

193 Appendix 2

Chem okine A ccession Chemokines Accession R eceptors Number Number CCR1 L10918 CXCLl P09341 CCR2 D29984 CXCL2 M36820 CCR3 U28694 CXCL3 M36821 CCR4 X85740 CXCL4 M25897 OCRS U54994 CXCL5 X78686 CCR6 U45984 CXCL6 U83303 CCR7 L31584 CXCL7 M54995 CCR8 U45983 CXCL8 P10145 CCR9 U45982 CXCL9 Q07325 CCR10 U13667 CXCLl 0 P02778 CCR11 API93507 CXCLl 1 U66096 CXCR1 U11870 CXCL12 U16752 CXCR2 LI9593 CXCLl 3 AF044196 CXCR3 X95876 CXCL14 AF073957 CXCR4 AF005058 CCL1 P22362 CXCR5 NM001716 CCL2 PI 3500 CX3CR1 P49238 CCL3 M23452 CCXCR1 L36149 SCYA3L M23178 D6 Y12815 CCL4 P I3236 DARC U01839 CCL5 P13501 CCL7 P80098 7-TM Receptors Accession CCL8 P80075 Number P2Y2 U07225 CCL11 P51671 PAR1 M62424 CCL13 U46767 PAR2 U34038 CCL14 AF031587 PAR3 U92971 CCL15 NM 004167 PAR4 AF080214 CCL16 ABO18249 C5aR M62505 CCL17 NM 002987 C3aR U28488 CCL18 AB012113 Edg1 M31210 CCL19 Q99731 Edg2 U78192 CCL20 P78556 Edg 3 X83864 CCL21 U88320 Edg4 AF011466 CCL22 U83171 Edg 5 AF034780 CCL23 U85767 Edg 6 NM 003775 CCL24 NM 002991 Edg 7 NM 012152 CCL25 015444 Edg 8 AC011461 CCL26 ABO10447 CB1 U73304 CCL27 AJ243542 CB2 X74328 CCL28 AF220210 EMR1 X81479 XCL1 NP 002986 CD97 U76764 CX3CL1 U91835

194 7-TM R eceptor A ccession Housekeeping Genes A ccession Ligands Num ber Num ber Thrombin J00307 alpha-catenin T65118 Trypsin 1 M22612 platelet-type phosphofructokinase R38433 Trypsin 2 M27602 lysosomal giycosyiasparaginase R42153 Trypsin 3 XI5505 ETS2 oncogene R42479 Trypsin 4 X72781 capping protein alpha R42609 Tryptase a M30038 mitochondrial short-chain enoyl-CoA R43558 hydratase Tryptase b M37488 cytoplasmic beta-actin Ml0277 Elastase 2A M16631 Unknown est hits R20593 Elastase 2B Ml6653 mitochondrial ADP/ADT translocator R53942 Neutrophil Elastase M34379 superoxide dismutase R52548 Elastase 3A D00306 protein tyrosine kinase (Syk) R59598 Elastase 3B NM 007352 Lipoamide dehydrogenase (NAD) L03490 Caldecrin (elastase S82198 Unknown - 2 0 est only N48956 IV) Neuromedin U X76029 mRNA for lactate dehydrogenase-A H05914 glutamate receptor 2 H06193 7-TM O rphan A ccession alpha- 2-macroglobulin H06516 R eceptors Number FPRL1 M84562 glyceraldehyde 3-phosphate dehydrogenase HI6958 G2A AF083955 deoxyhypusine synthase R37766 P2Y5 AF000546 eukaryotic initiation factor 2B-epsilon R54818 CML2 U77827 90 kD heat shock protein R44334 0GR1 U35398 histone H2B.1, 3' end H06295 GPR 18 L42324 ADP-ribosylation factor 1 H I1049 H963 AF002986 adenylyl cyclase-associated protein R37953 TDAG8 NM 003608 malate dehydrogenase R15814 C3AR AAC50374 chaperonin protein (Tcp20) H07880 P2Y9 NM 005296 liver glutamate dehydrogenase R54424 EBI2 L08177 DNA repair helicase (ERCC3) H20856 HSGPCP Y13583 endothelin -1 H11003 GPR105 D13626 DNA (cytosin-5)-methyltransferase H09055 GPR34 AF039686 uroporphyrinogen III synthase T82469 HS0GPCR1 NM 006056 mRNA for calmodulin R44288 GPR 38 AF034632 ribosomal protein L3 homologue H05820 GPR15 U34806 ADP/ATP translocase R61295 GPR31 U65402 cytochrome bc-1 complex core protein II R12802 GPR32 AF045764 nuclear ribonucleoprotein particle (hnRNP) C H05899 CRTH2 AB008535 Beta Actin 5' Fragment M l0277 GPR44 AF118265 Beta Actin middle Fragment Ml0277 GPR16 X98356 Beta Actin 3' Fragment Ml0277 GPR115 AJ249248 DAD-1 D15057 GPR128 D10923 SLUG-1 NM003068 GPR132 AA827664 CNR2 NM 001841 PTAFR D10202 CCRL2 U97123 CMKLR1 U79527 hm74 D10923 RDC1 U67784 TYMSTR U73531

Genes represented on the DNA microarray with GenBank accession numbers

195 Reference List

Abbracchio,M.P., BoeynaemsJ.M., Barnard,E.A., Boyer,J.L., Kennedy,C., Miras-Portugal,M.T., King,B.F., Gachet,C., Jacobson,K.A., Weisman,G.A., and Bumstock,G. (2003). Characterization of the UDP-glucose receptor (re-named here the P2Y(14) receptor) adds diversity to the P2Y receptor family. Trends Pharmacol. Sci. 24, 52-55.

Abe,M., Hiura,K., Wilde,J., Moriyama,K., Hashimoto,T., Ozaki,S., Wakatsuki,S., Kosaka,M., Kido,S., Inoue,D., and Matsumoto,T. (2002). Role for macrophage inflammatory protein (MIP)- 1 alpha and MIP-lbeta in the development of osteolytic lesions in multiple myeloma. Blood 100, 2195-2202.

Adema,G.J., Hartgers,F., Verstraten,R., de Vries,E., Marland,G., Menon,S., Foster,!., Xu,Y., Nooyen,P., McClanahan,T., Bacon,K.B., and Figdor,C.G. (1997). A dendritic-cell-derived C-C chemokine that preferentially attracts naive T cells. Nature 287, 713-717.

Ahamed,!. and Ali,H. (2002). Distinct roles of receptor phosphorylation, G protein usage, and mitogen-activated protein kinase activation on platelet activating factor-induced leukotriene C(4) generation and chemokine production. !. Biol. Chem. 277, 22685-22691.

Ahuja,S.K., Lee,!.C., and Murphy,P.M. (1996). CXC chemokines bind to unique sets of selectivity determinants that can function independently and are broadly distributed on multiple domains of human interleukin-8 receptor B. Determinants of high affinity binding and receptor activation are distinct. !. Biol. Chem. 271, 225-232.

Akahoshi,T., Sasahara,T., Namai,R., Matsui,T., Watabe,H., Kitasato,H., Inoue,M., and Kondo,H. (2003). Production of macrophage inflammatory protein 3alpha (MIP-3alpha) (CCL20) and MIP- 3beta (CCL19) by human peripheral blood neutrophils in response to microbial pathogens. Infect. Immun. 71, 524-526.

Akatsu,T., Tamura,T., Takahashi,N., Udagawa,N., Tanaka,S., Sasaki,T., Yamaguchi,A., Nagata,N., and Suda,T. (1992). Preparation and characterization of a mouse osteoclast-like multinucleated cell population. !. Bone Miner. Res. 7, 1297-1306.

Alon,U., Barkai,N., Notterman,D.A., Gish,K., Ybarra,S., Mack,D., and Levine,A.!. (1999). Broad patterns of gene expression revealed by clustering analysis of tumor and normal colon tissues probed by oligonucleotide arrays. Proc. Natl. Acad. Sci. U. S. A. 96, 6745-6750.

Ancuta,P., Rao,R., Moses,A., Mehle,A., Shaw,S.K., Luscinskas,F.W., and Gabuzda,D. (2003). Fractalkine preferentially mediates arrest and migration of CD 16+ monocytes. !. Exp. Med. 197, 1701-1707.

Arai,F., Miyamoto,T., 0hneda,0., Inada,T., Sudo,T., Brasel,K., Miyata,T., Anderson,D.M., and Suda,T. (1999). Commitment and differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor kappaB (RANK) receptors. !. Exp. Med. 190, 1741-1754.

Ardavin,C., Martinez,d.H., Martin,P., Anjuere,F., Arias,C.F., Marin,A.R., Ruiz,S., Parrillas,V., and Hemandez,H. (2001). Origin and differentiation of dendritic cells. Trends Immunol. 22, 691-700.

196 Arenberg,D.A., Keane,M.P., DiGiovine,B., Kunkel,S.L,, Morris,S.B., Xue,Y.Y., Burdick,M.D., Glass,M.C., Iannettoni,M.D., and Strieter,R.M. (1998). Epithelial-neutrophil activating peptide (ENA-78) is an important angiogenic factor in non-small cell lung cancer. J. Clin. Invest. 102, 465- 472.

Amett,T.R., Gibbons,D.C., Utting,J.C., Orriss,I.R., Hoebertz,A., Rosendaal,M., and Meghji,S. (2003). Hypoxia is a major stimulator of osteoclast formation and bone resorption. J. Cell Physiol. 796, 2-8.

Austyn,J.M. (1998). Dendritic cells. Curr. Opin. Hematol. 5, 3-15.

Baba,M., Imai,T., Nishimura,M., Kakizaki,M., Takagi,S., Hieshima,K., Nomiyama,H., and Yoshie,0. (1997). Identification of CCR6, the specific receptor for a novel lymphocyte-directed CC chemokine LARC. J. Biol. Chem. 272, 14893-14898.

Baekkevold,E.S., Yamanaka,T., Palframan,R.T., Carlsen,H.S., Reinholt,F.P., von Andrian,U.H., Brandtzaeg,P., and Haraldsen,G. (2001). The CCR7 ligand elc (CCL19) is transcytosed in high endothelial venules and mediates T cell recruitment. J. Exp. Med. 193, 1105-1112.

Baggiolini,M., Dewald,B., and Moser,B. (1994). Interleukin-8 and related chemotactic cytokines— CXC and CC chemokines. Adv. Immunol. 55, 97-179.

Baggiolini,M., Dewald,B., and Moser,B. (1997). Human chemokines: an update. Annu. Rev. Immunol. 15, 675-705.

Banchereau,!., Briere,F., Caux,C., Davoust,!., Lebecque,S., Liu,Y.J., Pulendran,B., and Palucka,K. (2000). Immunobiology of dendritic cells. Annu. Rev. Immunol. 18, 767-811.

Banchereau,!. and Steinman,R.M. (1998). Dendritic cells and the control of immunity. Nature 392, 245-252.

Belperio,J.A., Keane,M.P., Arenberg,D.A., Addison,C.L., Ehlert,J.E., Burdick,M.D., and Strieter,R.M. (2000). CXC chemokines in angiogenesis. J. Leukoc. Biol. 6 8 ,1-8.

Bender,A., Sapp,M., Schuler,G., Steinman,R.M., and Bhardwaj,N. (1996). Improved methods for the generation of dendritic cells from nonproliferating progenitors in human blood. J. Immunol. Methods 196, 121-135.

Bendre,M.S., Montague,D.C., Peery,T., Akel,N.S., Gaddy,D., and Suva,L.J. (2003). Interleukin-8 stimulation of osteoclastogenesis and bone resorption is a mechanism for the increased osteolysis of metastatic bone disease. Bone 33, 28-37.

Berchtold,S., Ogilvie,A.L., Bogdan,C., Muhl-Zurbes,P., Ogilvie,A., Schuler,G., and Steinkasserer,A. (1999). Human monocyte derived dendritic cells express functional P2X and P2Y receptors as well as ecto-nucleotidases. FEBS Lett. 458, 424-428.

Bemhard,H., Disis,M.L., Heimfeld,S., Hand,S., Gralow,J.R., and Cheever,M.A. (1995). Generation of immunostimulatory dendritic cells from human CD34+ hematopoietic progenitor cells of the bone marrow and peripheral blood. Cancer Res. 55,1099-1104.

197 Bevilacqua,A., Ceriani,M.C., Capaccioli,S., and Nicolin,A. (2003). Post-transcriptional regulation of gene expression by degradation of messenger RNAs. J. Cell Physiol. 195, 356-372.

Bleul,C.C., Farzan,M., Choe,H., Parolin,C., Clark-Lewis,!., Sodroski,!., and Springer,!.A. (1996). The lymphocyte chemoattractant SDF-1 is a ligand for LESTR/fusin and blocks HIV-1 entry. Nature 382, 829-833.

Bochner,B.S., Bickel,C.A., Taylor,M.L., MacGlashan,D.W., Jr., Gray,P.W., Raport,C.J., and Godiska,R. (1999). Macrophage-derived chemokine induces human eosinophil chemotaxis in a CC chemokine receptor 3- and CC chemokine receptor 4-independent manner. J. Allergy Clin. Immunol. 103, 527-532.

Boissy,P., Saltel,F., Bouniol,C., Jurdic,P., and Machuca-Gayet,I. (2002). Transcriptional activity of nuclei in multinucleated osteoclasts and its modulation by calcitonin. Endocrinology 1 4 3 ,1913- 1921.

Brinckerhoff,L.H., Thompson,!.W., and Slingluff,C.L., Jr. (2000). Melanoma vaccines. Curr. Opin. Oncol. 12, 163-173.

Brivanlou,A.H. and Damell,J.E., Jr. (2002). Signal transduction and the control of gene expression. Science 295, 813-818.

Brossart,M.D. (2002). Dendritic cells in vaccination therapies of malignant diseases. Transfus. Apheresis. Sci. 2 7 ,183-186.

Bug,G., Aman,M.J., Tretter,T., Huber,C., and Peschel,C. (1998). Induction of macrophage- inflammatory protein 1 alpha (MIP-1 alpha) by interferon-alpha. Exp. Hematol. 2 6 ,117-123.

Bumstock,G. (1997). The past, present and future of purine nucleotides as signalling molecules. Neuropharmacology 56, 1127-1139.

Cameron,S.B., Nawijn,M.C., Kum,W.W., Savelkoul,H.F., and Chow,A.W. (2001). Regulation of helper T cell responses to staphylococcal superantigens. Eur. Cytokine Netw. 12, 210-222.

Camussi,G., Tetta,C., and Baglioni,C. (1990). The role of platelet-activating factor in inflammation. Clin. Immunol. Immunopathol. 57, 331-338.

Canalis,E. (2003). Mechanisms of glucocorticoid-induced osteoporosis. Curr. Opin. Rheumatol. 15, 454-457.

Caux,C., Ait-Yahia,S., Chemin,K., de Bouteiller,0., Dieu-Nosjean,M.C., Homey,B., Massacrier,C., Vanbervliet,B., Zlotnik,A., and Vicari,A. (2000). Dendritic cell biology and regulation of dendritic cell trafficking by chemokines. Springer Semin. Immunopathol. 22, 345-369.

Caux,C., Dezutter-Dambuyant,C., Schmitt,D., and Banchereau,J. (1992). GM-CSF and TNF-alpha cooperate in the generation of dendritic Langerhans cells. Nature 360, 258-261.

Caux,C., Vanbervliet,B., Massacrier,C., Durand,!, and Banchereau,J. (1996). Interleukin-3 cooperates with tumor necrosis factor alpha for the development of human dendritic/Langerhans cells from cord blood CD34+ hematopoietic progenitor cells. Blood 87, 2376-2385.

198 Cazzola,M. and Skoda,R.C. (2000). Translational pathophysiology: a novel molecular mechanism of human disease. Blood 95, 3280-3288.

Cecchini,M.G., Dominguez,M.G., Mocci,S., Wetterwald,A., Felix,R., Fleisch,H., Chisholm,0., Hofstetter,W., Pollard,!.W., and Stanley,E.R. (1994). Role of colony stimulating factor-1 in the establishment and regulation of tissue macrophages during postnatal development of the mouse. Development 120, 1357-1372.

Cella,M., Sallusto,F., and Lanzavecchia,A. (1997). Origin, maturation and antigen presenting function of dendritic cells. Curr. Opin. Immunol. 9, 10-16.

Chackalamannil,S. (2003). G-protein coupled receptor antagonists-1: protease activated receptor-1 (PAR-1) antagonists as novel cardiovascular therapeutic agents. Curr. Top. Med. Chem. 3, 1115- 1123.

Chambers,J.K., Macdonald,L.E., Sarau,H.M., Ames,R.S., Freeman,K., Foley,J.J., Zhu,Y., McLaughlin,M.M., Murdock,P., McMillan,!., Trill,!, Swift,A., Aiyar,N., Taylor,P., Vawter,!., Naheed,S., Szekeres,P., Hervieu,G., Scott,C., Watson,J.M., Murphy,A.J., Duzic,E., Klein,C., Bergsma,D.!, Wilson,S., and !ivi,G.P. (2000). A G protein-coupled receptor for UDP-glucose. J. Biol. Chem. 275, 10767-1077!

Chandrasekar,B., Melby,P.C., Sarau,H.M., Raveendran,M., Perla,R.P., Marelli-Berg,F.M., Dulin,N.O., and Singh,I.S. (2003). Chemokine-cytokine cross-talk. The E!R+ CXC chemokine !IX (CXC!5) amplifies a proinflammatory cytokine response via a phosphatidylinositol 3-kinase- NF-kappa B pathway. J. Biol. Chem. 278, 4675-4686.

Charlton,M.E., Williams,A.S., Fogliano,M., Sweetnam,P.M., and Duman,R.S. (1997). The isolation and characterization of a novel G protein-coupled receptor regulated by immunologic challenge. Brain Res. 764, 141-148.

Chen,W. and !i,Y.P. (1998). Generation of mouse osteoclastogenic cell lines immortalized with SV40 large T antigen. J. Bone Miner. Res. 13, 1112-1123.

Choi,G.S., Kang,!M., and !ee,M.G. (2000). Analysis of methods for the generation of dendritic cells from human peripheral blood monocytes. Yonsei Med. J. 41, 642-650.

Chun,J., Goetzl,E.J., Hla,T., Igarashi,Y., !ynch,K.R., Moolenaar,W., Pyne,S., and Tigyi,G. (2002). International Union of Pharmacology. XXXIV. !ysophospholipid receptor nomenclature. Pharmacol. Rev. 54, 265-269.

Ciesielski,C.J., Andreakos,!., Foxwell,B.M., and Feldmann,M. (2002). TNFalpha-induced macrophage chemokine secretion is more dependent on NF-kappaB expression than lipopolysaccharides-induced macrophage chemokine secretion. Eur. J. Immunol. 32, 2037-2045.

Clark-Eewis,!, Dewald,B., Geiser,T., Moser,B., and Baggiolini,M. (1993). Platelet factor 4 binds to interleukin 8 receptors and activates neutrophils when its N terminus is modified with Glu-!eu-Arg. Proc. Natl. Acad. Sci. U. S. A 90, 3574-3577.

Collin-Osdoby,P. (1994). Role of vascular endothelial cells in bone biology. J. Cell Biochem. 55, 304-309.

199 Collin-Osdoby,P., Rothe,L., Anderson,F,, Nelson,M., Maloney,W., and Osdoby,P. (2001). Receptor activator of NF-kappa B and osteoprotegerin expression by human micro vascular endothelial cells, regulation by inflammatory cytokines, and role in human osteoclastogenesis. J. Biol. Chem. 276, 20659-20672.

Cook,D.N. (1996). The role of MIP-1 alpha in inflammation and hematopoiesis. J. Leukoc. Biol. 59, 61-66.

Denlinger,L.C., Fisette,P.L., Sommer,J.A., Watters,J.J., Prabhu,U., Dubyak,G.R., Proctor,R.A., and Bertics,P.J. (2001). Cutting edge: the nucleotide receptor P2X7 contains multiple protein- and lipid- interaction motifs including a potential binding site for bacterial lipopolysaccharide. J. Immunol. 167, 1871-1876.

Dietz,A.B., Bulur,P.A., Knutson,G.J., Matasic,R., and Vuk-Pavlovic,S. (2000). Maturation of human monocyte-derived dendritic cells studied by microarray hybridization. Biochem. Biophys. Res. Commun. 275, 731-738.

Dixon,R.A., Kobilka,B.K., Strader,D.J., Benovic,J.L., Dohlman,H.G., Frielle,T., Bolanowski,M.A., Bennett,C.D., Rands,E., and Diehl,R.E. (1986). Cloning of the gene and cDNA for mammalian beta-adrenergic receptor and homology with rhodopsin. Nature 321, 75-79.

Domotor,E., Bartha,K., Machovich,R., and Adam-Vizi,V. (2002). Protease-activated receptor-2 (PAR-2) in brain microvascular endothelium and its regulation by plasmin and elastase. J. Neurochem. 80, 746-754.

Driscoll,K.E. (1994). Macrophage inflammatory proteins: biology and role in pulmonary inflammation. Exp. Lung Res. 20, 473-490.

Dugast,C., Gaudin,A., and Toujas,L. (1997). Generation of multinucleated giant cells by culture of monocyte-derived macrophages with IL-4. J. Leukoc. Biol. 61, 517-521.

Dzionek,A., Fuchs,A., Schmidt,P., Cremer,S., Zysk,M., Miltenyi,S., Buck,D.W., and Schmitz,]. (2000). BDCA-2, BDCA-3, and BDCA-4: three markers for distinct subsets of dendritic cells in human peripheral blood. J. Immunol. 165, 6037-6046.

Ebner,S., Ratzinger,G., Krosbacher,B., Schmuth,M., Weiss,A., Reider,D., Kroczek,R.A., Herold,M., Heufler,C., Fritsch,P., and Romani,N. (2001). Production of IL-12 by human monocyte- derived dendritic cells is optimal when the stimulus is given at the onset of maturation, and is further enhanced by IL-4. J. Immunol. 166, 633-641.

Eigenbrodt,E., Reinacher,M., Scheefers-Borchel,U., Scheefers,H., and Friis,R. (1992). Double role for pyruvate kinase type M2 in the expansion of phosphometabolite pools found in tumor cells. Crit. Rev. Oncog. 5, 91-115.

Everts, V., Aronson,D.C., and Beertsen,W. (1985). Phagocytosis of bone collagen by osteoclasts in two cases of pycnodysostosis. Calcif. Tissue Int. 37, 25-31.

Feng,Y., Broder,C.C., Kennedy,P.E., and Berger,E.A. (1996). HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor. Science 272, 872-877.

200 Ferrari,D., Chiozzi,P., Falzoni,S., Dal Susino,M., Melchiorri,L., Baricordi,O.R., and Di Virgilio,F. (1997). Extracellular ATP triggers IL-1 beta release by activating the purinergic P2Z receptor of human macrophages. J. Immunol. 159,1451-1458.

Ferrari,D., la Sala,A., Chiozzi,P., Morelli,A., Falzoni,S., Girolomoni,G., Idzko,M., Dichmann,S., Norgauer,]., and Di Virgilio,F. (2000). The P2 purinergic receptors of human dendritic cells: identification and coupling to cytokine release. FASEB J. 14, 2466-2476.

Fischer,F.R., Luo,Y., Luo,M., Santambrogio,L., and Dorf,M.E. (2001). RANTES-induced chemokine cascade in dendritic cells. J. Immunol. 167, 1637-1643.

Flanagan,A.M. and Massey,H.M. (2003). Generating human osteoclasts in vitro from bone marrow and peripheral blood. Methods Mol. Med. 80, 113-128.

Forssmann,U., Delgado,M.B., Uguccioni,M., Loetscher,P., Garotta,G., and Baggiolini,M. (1997). CKbeta8, a novel CC chemokine that predominantly acts on monocytes. FEBS Lett. 408, 211-216.

Forssmann,U., Magert,H.J., Adermann,K., Escher,S.E., and Forssmann,W.G. (2001). Hemofiltrate CC chemokines with unique biochemical properties: HCC- l/CCL14a and HCC-2/CCL15. J. Leukoc. Biol. 70, 357-366.

Freeman,K., Tsui,P., Moore,D., Emson,P.C., Vawter,L., Naheed,S., Lane,P., Bawagan,H., Herrity,N., Murphy,K., Sarau,H.M., Ames,R.S., Wilson,S., Livi,G.P., and Chambers,J.K. (2001). Cloning, pharmacology, and tissue distribution of G-protein-coupled receptor GPR105 (KIAAOOOl) rodent orthologs. Genomics 7 8 ,124-128.

Fueller,M., Wang,d.A., Tigyi,G., and Siess,W. (2003). Activation of human monocytic cells by lysophosphatidic acid and sphingosine-1 -phosphate. Cell Signal. 15, 367-375.

Fuller,K., Wong,B., Fox,S., Choi,Y., and Chambers,T.J. (1998). TRANCE is necessary and sufficient for osteoblast-mediated activation of bone resorption in osteoclasts. J. Exp. Med. 188, 997-1001.

Gallucci,S. and Matzinger,P. (2001). Danger signals: SOS to the immune system. Curr. Opin. Immunol. 13, 114-119.

Gao,J.L., Kuhns,D.B., Tiffany,H.L., McDermott,D., Li,X., Francke,U., and Murphy,P.M. (1993). Structure and functional expression of the human macrophage inflammatory protein 1 alpha/RANTES receptor. J. Exp. Med. 177, 1421-1427.

Geissmann,F., Jung,S., and Littman,D.R. (2003). Blood monocytes consist of two principal subsets with distinct migratory properties. Immunity. 19, 71-82.

Gether,U. and Kobilka,B.K. (1998). G protein-coupled receptors. II. Mechanism of agonist activation. J. Biol. Chem. 273, 17979-17982.

Goetzl,E.J., Graeler,M., Huang,M.C., and Shankar,G. (2002). Lysophospholipid growth factors and their g protein-coupled receptors in immunity, coronary artery disease, and cancer. Scientific World Journal. 2, 324-338.

201 Gordon,s., (1996a) Weir’s Handbook of Experimental Immunology, volume 4, 5* edition; The myeloid system, 84-95

Gordon,S., (1996b) The Oxford Textbook of Medicine, volume 1, 3"^^ edition; The mononuclear phagocyte system and tissue homeostasis, 153.1-153.9

Graler,M.H., Bemhardt,G., and Lipp,M. (1999). A lymphoid tissue-specific receptor, EDG6, with potential immune modulatory functions mediated by extracellular lysophospholipids. Curr. Top. Microbiol. Immunol. 246, 131-136.

Granucci,F., Andrews,D.M., Degli-Esposti,M.A., and Ricciardi-Castagnoli,P. (2002). IL-2 mediates adjuvant effect of dendritic cells. Trends Immunol. 2 3 ,169-171.

Granucci,F., Ferrero,E., Foti,M., Aggujaro,D., Vettoretto,K., and Ricciardi-Castagnoli,P. (1999). Early events in dendritic cell maturation induced by EPS. Microbes. Infect. 1, 1079-1084.

Granucci,F., Vizzardelli,C., Pavelka,N., Feau,S., Persico,M., Virzi,E., Rescigno,M., Moro,G., and Ricciardi-Castagnoli,P. (2001). Inducible IL-2 production by dendritic cells revealed by global gene expression analysis. Nat. Immunol. 2, 882-J

Greaves,D.R., Hakkinen,T., Lucas,A.D., Liddiard,K., Jones,E., Quinn,C.M., Senaratne,!., Green,F.R., Tyson,K., Boyle,J., Shanahan,C., Weissberg,P.L., Gordon,S., and Yla-Hertualla,S. (2001). Linked chromosome 16ql3 chemokines, macrophage-derived chemokine, ffactalkine, and thymus- and activation-regulated chemokine, are expressed in human atherosclerotic lesions. Arterioscler. Thromb. Vase. Biol. 21, 923-929.

Greaves,D.R., Wang,W., Dairaghi,D.J., Dieu,M.C., Saint-Vis,B., Franz-Bacon,K., Rossi,D., Caux,C., McClanahan,T., Gordon,S., Zlotnik,A., and Schall,T.J. (1997). CCR6, a CC chemokine receptor that interacts with macrophage inflammatory protein 3alpha and is highly expressed in human dendritic cells. J. Exp. Med. 186, 837-844.

Grigoriadis,A.E., Wang,Z.Q., Cecchini,M.G., Hofstetter,W., Felix,R., Fleisch,H.A., and Wagner,E.F. (1994). c-Fos: a key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science 266, 443-448.

Gutzmer,R., Langer,K., Lisewski,M., Mommert,S., Rieckbom,D., Rapp,A., and Werfel,T. (2002). Expression and function of histamine receptors 1 and 2 on human monocyte-derived dendritic cells. J. Allergy Clin. Immunol. 109, 524-531.

Han,K.H., Tangirala,R.K., Green,S.R., and Quehenberger,0, (1998). Chemokine receptor CCR2 expression and monocyte chemoattractant protein-1-mediated chemotaxis in human monocytes. A regulatory role for plasma LDL. Arterioscler. Thromb. Vase. Biol. 18, 1983-1991.

Harada,A., Sekido,N., Akahoshi,T., Wada,T., Mukaida,N., and Matsushima,K. (1994). Essential involvement of interleukin-8 (IL-8) in acute inflammation. J. Leukoc. Biol. 56, 559-564.

Haribabu,B., Zhelev,D.V., Pridgen,B.C., Richardson,R.M., Ali,H., and Snyderman,R. (1999). Chemoattractant receptors activate distinct pathways for chemotaxis and secretion. Role of G- protein usage. J. Biol. Chem. 274, 37087-37092.

202 Harvey,K., Siddiqui,R.A., Sliva,D., Garcia,J.G., and English,D. (2002). Serum factors involved in human microvascular endothelial cell morphogenesis. J. Lab. Clin. Med. 140, 188-198.

Hashimoto,S., Suzuki,!., Dong,H.Y., Nagai,S., Yamazaki,N., and Matsushima,K. (1999). Serial analysis of gene expression in human monocyte-derived dendritic cells. Blood 94, 845-852.

Hayden,J.M., Brachova,L., Higgins,K., Obermiller,L., Sevanian,A., Khandrika,S., and Reaven,P.D. (2002). Induction of monocyte differentiation and foam cell formation in vitro by 7-ketocholesterol. J. Lipid Res. 43, 26-35.

Heidemann,!., Ogawa,H., Dwinell,M.B., Rafiee,P., Maaser,C., Gockel,H.R., Otterson,M.F., Ota,D.M., Lugering,N., Domschke,W., and Binion,D.G. (2003). Angiogenic effects of interleukin 8 (CXCL8) in human intestinal microvascular endothelial cells are mediated by CXCR2. J. Biol. Chem. 278, 8508-8515.

Hentunen,T.A., Reddy,S.V., Boyce,B.F., Devlin,R., Park,H.R., Chung,H., Selander,K.S., Dallas,M., Kurihara,N., Galson,D.L., Goldring,S.R., Koop,B.A., Windle,J.J., and Roodman,G.D. (1998). Immortalization of osteoclast precursors by targeting Bel -XL and Simian virus 40 large T antigen to the osteoclast lineage in transgenic mice. J. Clin. Invest 102, 88-97.

Hieshima,K., Imai,T., Baba,M., Shoudai,K., Ishizuka,K., Nakagawa,T., Tsuruta,J., Takeya,M., Sakaki,Y., Takatsuki,K., Miura,R., Opdenakker,G., Van Damme,J., Yoshie,0., and Nomiyama,H. (1997). A novel human CC chemokine PARC that is most homologous to macrophage- inflammatory protein-1 alpha/LD78 alpha and chemotactic for T lymphocytes, but not for monocytes. J. Immunol. 1 59,1140-1149.

Hill,S.J., Ganellin,C.R., Timmerman,H., Schwartz,J.C., Shankley,N.P., Young,J.M., Schunack,W., Levi,R., and Haas,H.L. (1997). International Union of Pharmacology. XIII. Classification of histamine receptors. Pharmacol. Rev. 49,253-278.

Hollopeter,G., Jantzen,H.M., Vincent,D., Li,G., England,L., Ramakrishnan,V., Yang,R.B., Nurden,P., Nurden,A., Julius,D., and Conley,P.B. (2001). Identification of the platelet ADP receptor targeted by antithrombotic drugs. Nature 409, 202-207.

Holmes,W.E., Lee,J., Kuang,W.J., Rice,G.C., and Wood,W.I. (1991). Structure and functional expression of a human interleukin-8 receptor. Science 2 5 3 ,1278-1280.

Horton,M.A., Dorey,E.L., Nesbitt,S.A., Samanen,!., Ali,F.E., Stadel,J.M., Nichols,A., Greig,R., and Helfrich,M.H. (1993). Modulation of vitronectin receptor-mediated osteoclast adhesion by Arg- Gly-Asp peptide analogs: a structure-function analysis. J. Bone Miner. Res. 8, 239-247.

Horwood,N.J., Kartsogiannis,V., Quinn,J.M., Romas,E., Martin,T.J., and Gillespie,M.T. (1999). Activated T lymphocytes support osteoclast formation in vitro. Biochem. Biophys. Res. Commun. 265, 144-150.

Howard,A.D., McAllister,G., Feighner,S.D., Liu,Q., Nargund,R.P., Van der Ploeg,L.H., and Patchett,A.A. (2001). Orphan G-protein-coupled receptors and natural ligand discovery. Trends Pharmacol. Sci. 2 2 ,132-140.

203 Idzko,M., la Sala,A., Ferrari,D., Panther,E., Herouy,Y., Dichmann,S., Mockenhaupt,M., Di Virgilio,F., Girolomoni,G., and Norgauer,J. (2002). Expression and function of histamine receptors in human monocyte- derived dendritic cells. J. Allergy Clin. Immunol. 109, 839-846.

Inaba,K., Inaba,M., Romani,N., Aya,H., Deguchi,M., Ikehara,S., Muramatsu,S., and Steinman,R.M. (1992). Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor. J. Exp. Med. 176, 1693- 1702.

Ito,T., Noguchi,Y., Udaka,N., Kitamura,H., and Satoh,S. (1999). Glucose transporter expression in developing fetal lungs and lung neoplasms. Histol. Histopathol. 14, 895-904.

Jacobs,K.A., Collins-Racie,L:A., Colbert,M., Duckett,M., Golden-Fleet,M., Kelleher,K., Kriz,R., LaVallie,E.R., Merberg,D., Spaulding,V., Stover,J., Williamson,M.J., and McCoy,J.M. (1997). A genetic selection for isolating cDNAs encoding secreted proteins. Gene 198, 289-296.

Jin,J., Dasari,V.R., Sistare,F.D., and Kunapuli,S.P. (1998). Distribution ofP2Y receptor subtypes on haematopoietic cells. Br. J. Pharmacol. 123, 789-794.

Jonuleit,H., Schmitt,E., Schuler,G., Knop,J., and Enk,A.H. (2000). Induction of interleukin 10- producing, nonproliferating CD4(+) T cells with regulatory properties by repetitive stimulation with allogeneic immature human dendritic cells. J. Exp. Med. 192, 1213-1222.

Kartsogiannis,V., Zhou,H., Horwood,N.J., Thomas,R.J., Hards,D.K., Quinn,J.M., Niforas,P., Ng,K.W., Martin,T.J., and Gillespie,M.T. (1999). Localization of RANKL (receptor activator of NF kappa B ligand) mRNA and protein in skeletal and extraskeletal tissues. Bone 25, 525-534.

Katou,F., Ohtani,H., Nakayama,T., Ono,K., Matsushima,K., Saaristo,A., Nagura,H., Yoshie,0., and Motegi,K. (2001). Macrophage-derived chemokine (MDC/CCL22) and CCR4 are involved in the formation of T lymphocyte-dendritic cell clusters in human inflamed skin and secondary lymphoid tissue. Am. J. Pathol. 1 58,1263-1270.

Kaur,C., Hao,A.J., Wu,C.H., and Ling,E.A. (2001). Origin of microglia. Microsc. Res. Tech. 54, 2- 9.

Kawabata,A. and Kuroda,R. (2000). Protease-activated receptor (PAR), a novel family of G protein-coupled seven trans-membrane domain receptors: activation mechanisms and physiological roles. Jpn. J. Pharmacol. 8 2 , 171-174.

Kellerman,D., Evans,R., Mathews,D., and Shaffer,C. (2002). Inhaled P2Y2 receptor agonists as a treatment for patients with Cystic Fibrosis lung disease. Adv. Drug Deliv. Rev. 54, 1463-1474.

Kim,Y.M., Kim,Y.M., Lee,Y.M., Kim,H.S., Kim,J.D., Choi,Y., Kim,K.W., Lee,S.Y., and Kwon,Y.G. (2002). TNF-related activation-induced cytokine (TRANCE) induces angiogenesis through the activation of Src and phospholipase C (PLC) in human endothelial cells. J. Biol. Chem. 277, 6799-6805.

Kirchhoff,K., Weinmarm,0., Zwimer,J., Begemann,G., Gotze,0., Rapp,A., and Werfel,T. (2001). Detection of anaphylatoxin receptors on CD83+ dendritic cells derived from human skin. Immunology 103, 210-217.

204 Koch,A.E., Volin,M.V., Woods,J.M., Kunkel,S.L., Connors,M.A., Harlow,L.A., Woodmff,D.C., Burdick,M.D., and Strieter,R.M. (2001). Regulation of angiogenesis by the C-X-C chemokines interleukin-8 and epithelial neutrophil activating peptide 78 in the rheumatoid joint. Arthritis Rheum. 44, 31-40.

Kohl,J. (2001). Anaphylatoxins and infectious and non-infectious inflammatory diseases. Mol. Immunol. 55, 175-187.

Kong,Y.Y., Boyle,W.J., and Penninger,J.M. (1999a). Osteoprotegerin ligand: a common link between osteoclastogenesis, lymph node formation and lymphocyte development. Immunol. Cell Biol. 77, 188-193.

Kong,Y.Y., Feige,U., Sarosi,L, Bolon,B., Tafuri,A., Morony,S., Capparelli,C., Li,J., Elliott,R., McCabe,S., Wong,T., Campagnuolo,G., Moran,E., Bogoch,E.R., Van,G., Nguyen,L.T., Ohashi,P.S., Lacey,D.L., Fish,E., Boyle,W.J., and Penninger,J.M. (1999b). Activated T cells regulate bone loss and joint destruction in adjuvant arthritis through osteoprotegerin ligand. Nature 402, 304-309.

Kong,Y.Y., Yoshida,H., Sarosi,I., Tan,H.L., Timms,E., Capparelli,C., Morony,S., Oliveira-dos- Santos,A.J., Van,G., Itie,A., Khoo,W., Wakeham,A., Dunstan,C.R., Lacey,D.L., Mak,T.W., Boyle,W.J., and Penninger,J.M. (1999c). OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 397, 315-323.

Kunkel,E.J. and Butcher,E.C. (2002). Chemokines and the tissue-specific migration of lymphocytes. Immunity. 16, 1-4. la Sala,A., Sebastiani,S., Ferrari,D., Di Virgilio,F., Idzko,M., Norgauer,J., and Girolomoni,G. (2002). Dendritic cells exposed to extracellular adenosine triphosphate acquire the migratory properties of mature cells and show a reduced capacity to attract type 1 T lymphocytes. Blood 99, 1715-1722.

Lacey,D.L., Timms,E., Tan,H.L., Kelley,M.J., Dunstan,C.R., Burgess,!., Elliott,R., Colombero,A., Elliott,G., Scully,S., Hsu,H., Sullivan,!., Hawkins,N., Davy,E., Capparelli,C., Eli,A., Qian,Y.X., Kaufman,S., Sarosi,L, Shalhoub,V., Senaldi,G., Guo,J., Delaney,!., and Boyle,W.J. (1998). Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 93, 165-176.

Lassus,!., Salo,!., !iranek,W.A., Santavirta,S., Nevalainen,!., Matucci-Cerinic,M., Horak,P., and Konttinen,Y. (1998). Macrophage activation results in bone resorption. Clin. Orthop. 7-15.

Lazarowski,E.R., Shea,D.A., Boucher,R.C., and Harden,T.K. (2003). Release of cellular UDP- glucose as a potential extracellular signaling molecule. Mol. Pharmacol. 63, 1190-1197.

Le Naour,F., Hohenkirk,L., Grolleau,A., Misek,D.E., Lescure,P., Geiger,!.D., Hanash,S., and Beretta,L. (2001). Profiling changes in gene expression during differentiation and maturation of monocyte-derived dendritic cells using both oligonucleotide microarrays and proteomics. !. Biol. Chem. 276, 17920-17931.

Lean,!.M., Murphy,C., Fuller,K., and Chambers,!.!. (2002). CCL9/MIP-1 gamma and its receptor CCRl are the major chemokine ligand/receptor species expressed by osteoclasts. !. Cell Biochem. 57,386-393.

205 Lee,B.c., Cheng,T., Adams,G.B., Attar,E.C., Miura,N., Lee,S.B., Saito,Y., Olszak,!., Dombkowski,D., 01son,D.P., Hancock,!., Choi,P.S., Haber,D.A., Luster,A.D., and Scadden,D.T. (2003). P2Y-like receptor, GPR105 (P2Y14), identifies and mediates chemotaxis of bone-marrow hematopoietic stem cells. Genes Dev. 17, 1592-1604.

Lee,!., Horuk,R., Rice,G.C., Bennett,G.L., Camerato,!., and Wood,W.I. (1992). Characterization of two high affinity human interleukin-8 receptors. !. Biol. Chem. 267, 16283-16287.

Li,!., Sarosi,L, Yan,X.Q., Morony,S., Capparelli,C., Tan,H.L., McCabe,S., Elliott,R., Scully,S., Van,G., Kaufman,S., !uan,S.C., Sun,Y., Tarpley,!., Martin,L., Christensen,K., McCabe,!., Kostenuik,P., Hsu,H., Fletcher,F., Dunstan,C.R., Lacey,D.L., and Boyle,W.!. (2000). RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Natl. Acad. Sci. U. S. A 9 7 ,1566-1571.

Liang,P. and Pardee,A.B. (1998). Differential display. A general protocol. Mol. Biotechnol. 10, 261-267.

Liao,F., Shirakawa,A.K., Foley,!.F., Rabin,R.L., and Farber,!.M. (2002). Human B cells become highly responsive to macrophage-inflammatory protein-3 alpha/CC chemokine ligand-20 after cellular activation without changes in CCR6 expression or ligand binding. !. Immunol. 168, 4871- 4880.

Liu,L.M. and MacPherson,G.G. (1993). Antigen acquisition by dendritic cells: intestinal dendritic cells acquire antigen administered orally and can prime naive T cells in vivo. !. Exp. Med. 177, 1299-1307.

Lockhart,D.!. and Winzeler,E.A. (2000). Genomics, gene expression and DNA arrays. Nature 405, 827-836.

Loetscher,M., Gerber,B., Loetscher,P., !ones,S.A., Piali,L., Clark-Lewis,L, Baggiolini,M., and Moser,B. (1996). Chemokine receptor specific for IP 10 and mig: structure, function, and expression in activated T-lymphocytes. !. Exp. Med. 184, 963-969.

Lucas,A.D., Chadwick,N., Warren,B.F., !ewell,D.P., Gordon,S., Powrie,F., and Greaves,D.R. (2001). The transmembrane form of the CX3CL1 chemokine ffactalkine is expressed predominantly by epithelial cells in vivo. Am. !. Pathol. 158, 855-866.

Luther,S.A. and Cyster,!.G. (2001). Chemokines as regulators of T cell differentiation. Nat. Immunol. 2, 102-107.

MacDonald,B.R., Takahashi,N., McManus,L.M., Holahan,!., Mundy,G.R., and Roodman,G.D. (1987). Formation of multinucleated cells that respond to osteotropic hormones in long term human bone marrow cultures. Endocrinology 120, 2326-2333.

Mackensen,A., Herbst,B., Kohler,G., Wolff-Vorbeck,G., Rosenthal,F.M., Veelken,H., Kulmburg,P., Schaefer,H.E., Mertelsmann,R., and Lindemann,A. (1995). Delineation of the dendritic cell lineage by generating large numbers of Birbeck granule-positive Langerhans cells from human peripheral blood progenitor cells in vitro. Blood 86, 2699-2707.

206 Maione,T.E., Gray,G.S., Petro,J., Hunt,A.J., Donner,A.L., Bauer,S.L, Carson,H.F., and Sharpe,RJ. (1990). Inhibition of angiogenesis by recombinant human platelet factor-4 and related peptides. Science 247, 77-79.

Marks,S.C., Jr. and Lane,P.W. (1976). Osteopetrosis, a new recessive skeletal mutation on chromosome 12 of the mouse. J. Hered. 6 711-18. ,

Marrero,M.B., Schieffer,B., Paxton,W.G., Heerdt,L., Berk,B.C., Delafontaine,P., and Bemstein,K.E^(1995). Direct stimulation of Jak/STAT pathway by the angiotensin II ATI receptor. Nature 575, 247-250.

Massey,H.M. and Flanagan,A.M. (1999). Human osteoclasts derive from CD 14-positive monocytes. Br. J. Haematol. 106, 167-170.

Massey,H.M., Scopes,!., Horton,M.A., and Flanagan,A.M. (2001). Transforming growth factor- betal (TGF-beta) stimulates the osteoclast-forming potential of peripheral blood hematopoietic precursors in a lymphocyte-rich microenvironment. Bone 28, 577-582.

Matsuzaki,K., Udagawa,N., Takahashi,N., Yamaguchi,K., Yasuda,H., Shima,N., Morinaga,T., Toyama,Y., Yabe,Y., Higashio,K., and Suda,T. (1998). Osteoclast differentiation factor (ODF) induces osteoclast-like cell formation in human peripheral blood mononuclear cell cultures. Biochem. Biophys. Res. Commun. 24 6199-204. ,

Mazzoni,A., Young,H.A., Spitzer,!.H., Visintin,A., and Segal,D.M. (2001). Histamine regulates cytokine production in maturing dendritic cells, resulting in altered T cell polarization. !. Clin. Invest. 108, 1865-1873.

McGowan,N.W., Walker,E.!., Macpherson,H., Ralston,S.H., and Helfrich,M.H. (2001). Cytokine- activated endothelium recruits osteoclast precursors. Endocrinology 142, 1678-1681.

McKnight,A.!. and Gordon,S. (1998). The EGF-TM7 family; unusual structures at the leukocyte surface. !. Leukoc. Biol. 63, 271-280.

McLellan,A.D., Starling,G.C., and Hart,D.N. (1995a). Isolation of human blood dendritic cells by discontinuous Nycodenz gradient centrifugation. !. Immunol. Methods 184, 81-89.

McLellan,A.D., Starling,G.C., Williams,L.A., Hock,B.D., and Hart,D.N. (1995b). Activation of human peripheral blood dendritic cells induces the CD86 co-stimulatory molecule. Eur. !. Immunol. 25, 2064-2068.

Min,!.K., Kim,Y.M., Kim,Y.M., Kim,E.C., Gho,Y.S., Kang,I.!., Lee,S.Y., Kong,Y.Y., and Kwon,Y.G. (2003). Vascular endothelial growth factor up-regulates expression of receptor activator of NF-kB (RANK) in endothelial cells: Concomitant increase of angiogenic responses to RANK ligand. !. Biol. Chem. 278, 39548-39557

Miotto,D., Christodoulopoulos,P., 01ivenstein,R., Taha,R., Cameron,L., Tsicopoulos,A., Tonnel,A.B., Fahy,0., Lafitte,!.!., Luster,A.D., Wallaert,B., Mapp,C.E., and Hamid,Q. (2001). Expression of IFN-gamma-inducible protein; monocyte chemotactic proteins 1,3, and 4; and eotaxin in THl- and TH2-mediated lung diseases. !. Allergy Clin. Immunol. 107, 664-670.

207 Miyamoto,T., 0hneda,0., Arai,F., Iwamoto,K., Okada,S., Takagi,K., Anderson,D.M., and Suda,T. (2001). Bifurcation of osteoclasts and dendritic cells from common progenitors. Blood 98, 2544- 2554.

Morrison,S.J., Uchida,N., and Weissman,I.L. (1995). The biology of hematopoietic stem cells. Annu. Rev. Cell Dev. Biol. 11, 35-71.

Mukaida,N., Harada,A., and Matsushima,K. (1998). Interleukin-8 (IL-8) and monocyte chemotactic and activating factor (MCAF/MCP-1), chemokines essentially involved in inflammatory and immune reactions. Cytokine Growth Factor Rev. 9, 9-23.

Murphy,P.M. and Tiffany,H.L. (1991). Cloning of complementary DNA encoding a functional human interleukin-8 receptor. Science 253, 1280-1283.

Nakagawa,M., Kaneda,T., Arakawa,T., Morita,S., Sato,T., Yomada,T., Hanada,K., Kumegawa,M., and Hakeda,Y. (2000). Vascular endothelial growth factor (VEGF) directly enhances osteoclastic bone resorption and survival of mature osteoclasts. FEBS Lett. 473, 161-164.

Nakao,M., Nomiyama,H., and Shimada,K. (1990). Structures of human genes coding for cytokine LD78 and their expression. Mol. Cell Biol. 10, 3646-3658.

Napolitano,M. and Santoni,A. (1999). Structure and function of the CC chemokine receptor (CCR) 8. Forum (Genova. ) 9, 315-324.

Nardelli,B., Morahan,D.K., Bong,G.W., Semenuk,M.A., Kreider,B.L., and Garotta,G. (1999a). Dendritic cells and MPIF-1: chemotactic activity and inhibition of endogenous chemokine production by IFN-gamma and CD40 ligation. J. Leukoc. Biol. 65, 822-828.

Nardelli,B., Tiffany,H.L., Bong,G.W., Yourey,P.A., Morahan,D.K., Li,Y., Murphy,P.M., and Alderson,R.F. (1999b). Characterization of the signal transduction pathway activated in human monocytes and dendritic cells by MPIF-1, a specific ligand for CC chemokine receptor 1. J. Immunol. 162, 435-444.

Naruse,K., Ueno,M., Satoh,T., Nomiyama,H., Tei,H., Takeda,M., Ledbetter,D.H., Coillie,E.V., Opdenakker,G., Gunge,N., Sakaki,Y., Iio,M., and Miura,R. (1996). A YAC contig of the human CC chemokine genes clustered on chromosome 17qll.2. Genomics 34, 236-240.

Nathans,!, and Hogness,D.S. (1983). Isolation, sequence analysis, and intron-exon arrangement of the gene encoding bovine rhodopsin. Cell 34, 807-814.

Nelson,P.!. and Krensky,A.M. (2001). Chemokines, chemokine receptors, and allograft rejection. Immunity. 14, 377-386.

Neote,K., DiGregorio,D., Mak,!.Y., Horuk,R., and Schall,T.!. (1993). Molecular cloning, functional expression, and signaling characteristics of a C-C chemokine receptor. Cell 72, 415-425.

Neufeld,G., Cohen,T., Gengrinovitch,S., and Poltorak,Z. (1999). Vascular endothelial growth factor (VEGF) and its receptors. FASEB !. 13, 9-22.

Ni,K. and 0'Neill,H.C. (1997). The role of dendritic cells in T cell activation. Immunol. Cell Biol. 75, 223-230.

208 NibbSjRJ., Yang,J,, Landau,N.R., Mao,J.H., and Graham,GJ. (1999). LD78beta, a non-allelic variant of human MIP-1 alpha (LD78alpha), has enhanced receptor interactions and potent HIV suppressive activity. J. Biol. Chem. 2 7 4 ,17478-17483.

Nicholson,G.C., Malakellis,M., Collier,F.M., Cameron,P.U., Holloway,W.R., Gough,T.J., Gregorio-King,C., Kirkland,M.A., and Myers,D.E. (2000). Induction of osteoclasts from CD 14- positive human peripheral blood mononuclear cells by receptor activator of nuclear factor kappaB ligand (RANKL). Clin. Sci. (Colch. ) 99, 133-140.

Niida,S., Kaku,M., Amano,H., Yoshida,H., Kataoka,H., Nishikawa,S., Tanne,K., Maeda,N., Nishikawa,S., and Kodama,H. (1999). Vascular endothelial growth factor can substitute for macrophage colony-stimulating factor in the support of osteoclastic bone resorption. J. Exp. Med. 190, 293-298.

Nomura,H., Nielsen,B.W., and Matsushima,K. (1993). Molecular cloning of cDNAs encoding a LD78 receptor and putative leukocyte chemotactic peptide receptors. Int. Immunol. 5, 1239-1249.

Nomura,N., Miyajima,N., Sazuka,T., Tanaka,A., Kawarabayasi,Y., Sato,S., Nagase,T., Seki,N., Ishikawa,K., and Tabata,S. (1994). Prediction of the coding sequences of unidentified human genes. I. The coding sequences of 40 new genes (KIAAOOOl-KIAA0040) deduced by analysis of randomly sampled cDNA clones from human immature myeloid cell line KG-1 (supplement). DNA Res. 1, 47-56.

North,R.A. (2002). Molecular physiology of P2X receptors. Physiol. Rev. 82, 1013-1067.

0'Keeffe,M., Hochrein,H., Vremec,D., Scott,B., Hertzog,P., Tatarczuch,L., and Shortman,K. (2003). Dendritic cell precursor populations of mouse blood: identification of the murine homologues of human blood plasmacytoid pre-DC2 and GDI lc+ DCl precursors. Blood 101, 1453-1459.

Oberlin,E., Amara,A., Bachelerie,F., Bessia,C., Virelizier,J.L., Arenzana-Seisdedos,F., Schwartz,0., Heard,J.M., Clark-Lewis,L, Legler,D.F., Loetscher,M., Baggiolini,M., and Moser,B. (1996). The CXC chemokine SDF-1 is the ligand for LESTR/fusin and prevents infection by T-cell- line-adapted HIV-l. Nature 382, 833-835.

Osdoby,P., Martini,M.C., and Caplan,A.I. (1982). Isolated osteoclasts and their presumed progenitor cells, the monocyte, in culture. J. Exp. Zool. 224, 331-344.

Palczewski,K., Kumasaka,T., Hori,T., Behnke,C.A., Motoshima,H., Fox,B.A., Le,T., I, Teller,D.C., Okada,T., Stenkamp,R.E., Yamamoto,M., and Miyano,M. (2000). Crystal structure of rhodopsin: A G protein-coupled receptor. Science 289, 739-745.

Paleolog,E.M. (1996). Angiogenesis: a critical process in the pathogenesis of RA-a role for VEGF? Br. J. Rheumatol. 35, 917-919.

Palucka,K.A., Taquet,N., Sanchez-Chapuis,F., and Gluckman,J.C. (1998). Dendritic cells as the terminal stage of monocyte differentiation. J. Immunol. 160, 4587-4595.

Pashenkov,M., Teleshova,N., Kouwenhoven,M., Kostulas,V., Huang,Y.M., Soderstrom,M., and Link,H. (2002). Elevated expression of CCR5 by myeloid (GDI lc+) blood dendritic cells in multiple sclerosis and acute optic neuritis. Glin. Exp. Immunol. 127, 519-526.

209 Pérou,C.M., Jeffrey,S.S., van de,R.M., Rees,C.A., Eisen,M.B., Ross,D.T., Pergamenschikov,A., Williams,C.F., Zhu,S.X., Lee,J.C., Lashkari,D., Shalon,D., Brown,P.O., and Botstein,D. (1999). Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc. Natl. Acad. Sci. U. S. A 96, 9212-9217.

Peters,J.H., Ruhl,S., and Friedrichs,D. (1987). Veiled accessory cells deduced from monocytes. Immunobiology 7 76, 154-166.

Pierce,K.L., Premont,R.T., and Lefkowitz,R.J. (2002). Seven-transmembrane receptors. Nat. Rev. Mol. Cell Biol. 3, 639-650.

Power,C.A., Church,D.J., Meyer,A., Alouani,S., Proudfoot,A.E., Clark-Lewis,L, Sozzani,S., Mantovani,A., and Wells,T.N. (1997). Cloning and characterization of a specific receptor for the novel CC chemokine MIP-3alpha from lung dendritic cells. J. Exp. Med. 186, 825-835.

Proctor,R.A., Denlinger,L.C., Leventhal,P.S., Daugherty,S.K., van de Loo,J.W., Tanke,T., Firestein,G.S., and Bertics,P.J. (1994). Protection of mice from endotoxic death by 2-methylthio- ATP. Proc. Natl. Acad. Sci. U. S. A 91, 6017-6020.

Quinn, J.M., Elliott,!., Gillespie,M.T., and Martin,!.!. (1998). A combination of osteoclast differentiation factor and macrophage-colony stimulating factor is sufficient for both human and mouse osteoclast formation in vitro. Endocrinology 139, 4424-4427.

Reddy,A., Sapp,M., Feldman,M., Subklewe,M., and Bhardwaj,N. (1997). A monocyte conditioned medium is more effective than defined cytokines in mediating the terminal maturation of human dendritic cells. Blood 90, 3640-3646.

Rescigno,M., Granucci,F., Citterio,S., Foti,M., and Ricciardi-Castagnoli,P. (1999). Coordinated events during bacteria-induced DC maturation. Immunol. Today 20, 200-203.

Rescigno,M., Winzler,C., Delia,D., Mutini,C., Lutz,M., and Ricciardi-Castagnoli,P. (1997). Dendritic cell maturation is required for initiation of the immune response. !. Leukoc. Biol, di, 415- 421.

Roglic,A., Prossnitz,E.R., Cavanagh,S.L., Pan,Z., Zou,A., and Ye,R.D. (1996). cDNA cloning of a novel G protein-coupled receptor with a large extracellular loop structure. Biochim. Biophys. Acta. 1305, 39-43.

Romagnani,S. (1999). Thl/Th2 cells. Inflamm. Bowel. Dis. 5, 285-294.

Ross,D.T., Scherf,U., Eisen,M.B., Perou,C.M., Rees,C., Spellman,P., Iyer,V., !effrey,S.S., van de,R.M., Waltham,M., Pergamenschikov,A., Lee,!.C., Lashkari,D., Shalon,D., Myers,T.G., Weinstein,!.N., Botstein,D., and Brown,P.O. (2000). Systematic variation in gene expression patterns in human cancer cell lines. Nat. Genet. 24, 227-235.

Rossi,D. and Zlotnik,A. (2000). The biology of chemokines and their receptors. Annu. Rev. Immunol. 18, 217-242.

Rothe,L., Collin-Osdoby,P., Chen,Y., Sunyer,T., Chaudhary,L., Tsay,A., Goldring,S., Avioli,L., and Osdoby,P. (1998). Human osteoclasts and osteoclast-like cells synthesize and release high basal

210 and inflammatory stimulated levels of the potent chemokine interleukin-8. Endocrinology 139, 4353-4363.

Ruffing,N., Sullivan,N., Sharmeen,L., Sodroski,!., and Wu,L. (1998). CCR5 has an expanded ligand-binding repertoire and is the primary receptor used by MCP-2 on activated T cells. Cell Immunol. 189, 160-168.

Sallusto,F. and Lanzavecchia,A. (1994). Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and downregulated by tumor necrosis factor alpha. J. Exp. Med. 179,1109-1118.

Sallusto,F. and Lanzavecchia,A. (2000). Understanding dendritic cell and T-lymphocyte traffic through the analysis of chemokine receptor expression. Immunol. Rev. 177,134-140.

Sallusto,F., Palermo,B., Lenig,D., Miettinen,M., Matikainen,S., Julkunen,L, Forster,R., Burgstahler,R., Lipp,M., and Lanzavecchia,A. (1999). Distinct patterns and kinetics of chemokine production regulate dendritic cell function. Eur. J. Immunol. 29, 1617-1625.

Sallusto,F., Schaerli,P., Loetscher,P., Schaniel,C., Lenig,D., Mackay,C.R., Qin,S., and Lanzavecchia,A. (1998). Rapid and coordinated switch in chemokine receptor expression during dendritic cell maturation. Eur. J. Immunol. 28, 2760-2769.

Sauter,B., Albert,M.L., Francisco,L., Larsson,M., Somersan,S., and Bhardwaj,N. (2000). Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells. J. Exp. Med. 191, 423-434.

Schaniel,C., Sallusto,F., Ruedl,C., Sideras,P., Melchers,F., and Rolink,A.G. (1999). Three chemokines with potential functions in T lymphocyte-independent and -dependent B lymphocyte stimulation. Eur. J. Immunol. 29, 2934-2947.

Scheicher,C., Mehlig,M., Zecher,R., and Reske,K. (1992). Dendritic cells from mouse bone marrow: in vitro differentiation using low doses of recombinant granulocyte-macrophage colony- stimulating factor. J. Immunol. Methods 154, 253-264.

Schnurr,M., Then,F., Galambos,P., Scholz,C., Siegmund,B., Endres,S., and Eigler,A. (2000). Extracellular ATP and TNF-alpha synergize in the activation and maturation of human dendritic cells. J. Immunol. 165, 4704-4709.

Selander,K., Lehenkari,P., and Vaananen,H.K. (1994). The effects of bisphosphonates on the resorption cycle of isolated osteoclasts. Calcif. Tissue Int. 55, 368-375.

Servet-Delprat,C., Amaud,S., Jurdic,P., Nataf,S., Grasset,M.F., Soulas,C., Domenget,C., Destaing,0., Rivollier,A., Perret,M., Dumontel,C., Hanau,D., Gilmore,G.L., Belin,M.F., Rabourdin- Combe,C., and Mouchiroud,G. (2002). Flt3+ macrophage precursors commit sequentially to osteoclasts, dendritic cells and microglia. BMC. Immunol. 5, 15.

Shalhoub,V., Elliott,G., Chiu,L., Manoukian,R., Kelley,M., Hawkins,N., Davy,E., Shimamoto,G., Beck,!., Kaufinan,S.A., Van,G., Scully,S., Qi,M., Grisanti,M., Dunstan,C., Boyle,W.!., and Lacey,D.L. (2000). Characterization of osteoclast precursors in human blood. Br. !. Haematol. Ill, 501-512.

211 Shortman,K. and Caux,C. (1997). Dendritic cell development: multiple pathways to nature's adjuvants. Stem Cells 15, 409-419.

Shortman,K. and Liu,Y.J. (2002). Mouse and human dendritic cell subtypes. Nat. Rev. Immunol. 2, 151-161.

Simonet,W.S., Lacey,D.L., Dunstan,C.R., Kelley,M., Chang,M.S., Luthy,R., Nguyen,H.Q., Wooden,S., Bennett,L., Boone,T., Shimamoto,G., DeRose,M., Elliott,R., Colombero,A., Tan,H.L., Trail,G., Sullivan,!., Davy,E., Bucay,N., Renshaw-Gegg,L., Hughes,T.M., Hill,D., Pattison,W., Campbell,P., Sander,S., Van,G., Tarpley,!., Derby,P., Lee,R and Boyle,W.!. (1997). Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89, 309- 319.

Sozzani,S., Allavena,P., D'Amico,G., Luini,W., Bianchi,G., Kataura,M., Imai,T., Yoshie,0., Bonecchi,R., and Mantovani,A. (1998). Differential regulation of chemokine receptors during dendritic cell maturation: a model for their trafficking properties. !. Immunol. 161, 1083-1086.

Sozzani,S., Luini,W., Borsatti,A., Polentarutti,N., Zhou,D., Piemonti,L., D'Amico,G., Power,C.A., Wells,T.N., Gobbi,M., Allavena,P., and Mantovani,A. (1997). Receptor expression and responsiveness of human dendritic cells to a defined set of CC and CXC chemokines. !. Immunol. 159, 1993-2000.

Spellberg,B. and Edwards,!.E., !r. (2001). Type 1/Type 2 immunity in infectious diseases. Clin. Infect. Dis. 32, 76-102.

Spisek,R., Bretaudeau,L., Barbieux,L, Meflah,K., and Gregoire,M. (2001). Standardized generation of fully mature p70 IL-12 secreting monocyte- derived dendritic cells for clinical use. Cancer Immunol. Immunother. 50,417-427.

Steinman,R.M., Pack,M., and Inaba,K. (1997). Dendritic cells in the T-cell areas of lymphoid organs. Immunol. Rev. 156, 25-37.

Strieter,R.M., Polverini,P.!., Kunkel,S.L., Arenberg,D.A., Burdick,M.D., Kasper,!., Dzuiba,!., Van Damme,!., Walz,A., Marriott,D., Chan, S.Y., Roczniak,S. and Shanafelt,A.B and. (1995). The fimctional role of the ELR motif in CXC chemokine-mediated angiogenesis. !. Biol. Chem. 270, 27348-27357.

Strobl,H., Riedl,E., Scheinecker,C., Bello-Femandez,C., Pickl,W.F., Rappersberger,K., Majdic,0., and Knapp,W. (1996). TGF-beta 1 promotes in vitro development of dendritic cells from CD34+ hemopoietic progenitors. !. Immunol. 157, 1499-1507.

Suda,T., Takahashi,N., and Martin,T.!. (1992). Modulation of osteoclast differentiation. Endocr. Rev. 13, 66-80.

Taha,R.A., Minshall,E.M., Leung,D.Y., Boguniewicz,M., Luster,A., Muro,S., Toda,M., and Hamid,Q.A. (2000). Evidence for increased expression of eotaxin and monocyte chemotactic protein-4 in atopic dermatitis. !. Allergy Clin. Immunol. 105, 1002-1007.

Takahashi,N., Yamana,H., Yoshiki,S., Roodman,G.D., Mundy,G.R., !ones,S.!., Boyde,A., and Suda,T. (1988). Osteoclast-like cell formation and its regulation by osteotropic hormones in mouse bone marrow cultures. Endocrinology 122, 1373-1382.

212 Takahashi,S,, Goldring,S., Katz,M., Hilsenbeck,S., Williams,R., and Roodman,G.D. (1995). Downregulation of calcitonin receptor mRNA expression by calcitonin during human osteoclast­ like cell differentiation. J. Clin. Invest. 9 5 ,167-171.

Tezuka,K., Nemoto,K., Tezuka,Y., Sato,T., Ikeda,Y., Kobori,M., Kawashima,H., Eguchi,H., Hakeda,Y., and Kumegawa,M. (1994). Identification of matrix metalloproteinase 9 in rabbit osteoclasts. J. Biol. Chem. 269, 15006-15009.

Thavarajah,M., Evans,D.B., and Kanis,J.A. (1991). l,25(OH)2D3 induces differentiation of osteoclast-like cells from human bone marrow cultures. Biochem. Biophys. Res. Commun. 176, 1189-1195.

Tonetti,M., Sturla,L., Giovine,M., Benatti,U., and De Flora,A. (1995). Extracellular ATP enhances mRNA levels of nitric oxide synthase and TNF- alpha in lipopolysaccharide-treated RAW 264.7 murine macrophages. Biochem. Biophys. Res. Commun. 214, 125-130.

Triantafilou,M. and Triantafilou,K. (2002). Lipopolysaccharide recognition: CD 14, TLRs and the LPS-activation cluster. Trends Immunol. 23, 301-304.

Tunaru,S., Kero,J., Schaub,A., Wufka,C., Blaukat,A., Pfeffer,K., and Offermanns,S. (2003). PUMA-G and HM74 are receptors for nicotinic acid and mediate its anti- lipolytic effect. Nat. Med. P, 352-355.

Udagawa,N., Takahashi,N., Akatsu,T., Sasaki,T., Yamaguchi,A., Kodama,H., Martin,T.J., and Suda,T. (1989). The bone marrow-derived stromal cell lines MC3T3-G2/PA6 and ST2 support osteoclast-like cell differentiation in cocultures with mouse spleen cells. Endocrinology 125, 1805- 1813.

Unanue,E.R. (1984). Antigen-presenting function of the macrophage. Annu. Rev. Immunol. 2, 395- 428.

Veikkola,T. and Alitalo,K. (1999). VEGFs, receptors and angiogenesis. Semin. Cancer Biol. 9, 211- 220.

Verhasselt,V., Buelens,C., Willems,F., De Groote,D., Haeffher-Cavaillon,N., and Goldman,M. (1997). Bacterial lipopolysaccharide stimulates the production of cytokines and the expression of costimulatory molecules by human peripheral blood dendritic cells: evidence for a soluble CD 14- dependent pathway. J. Immunol. 158, 2919-2925.

Vissers,J.L., Hartgers,F.C., Lindhout,E., Teunissen,M.B., Figdor,C.G., and Adema,G.J. (2001). Quantitative analysis of chemokine expression by dendritic cell subsets in vitro and in vivo. J Leukoc. Biol. 69, 785-793.

Votta,B.J., White,J.R., Dodds,R.A., James,I.E., Connor,J.R., Lee-Rykaczewski,E., Eichman,C.F., Kumar,S., Lark,M.W., and Gowen,M. (2000). CKbeta-8 [CCL23], a novel CC chemokine, is chemotactic for human osteoclast precursors and is expressed in bone tissues. J. Cell Physiol. 183, 196-207.

Vulcano,M., Albanesi,C., Stoppacciaro,A., Bagnati,R., D'Amico,G., Struyf,S., Transidico,P., Bonecchi,R., Del Prete,A., Allavena,P., Ruco,L.P., Chiabrando,C., Girolomoni,G., Mantovani,A.,

213 and Sozzani,S. (2001). Dendritic cells as a major source of macrophage-derived chemokine/CCL22 in vitro and in vivo. Eur. J. Immunol. 31, 812-822.

VulcanOjM., Struyf,S., Scapini,P., Cassatella,M., Bemasconi,S., Bonecchi,R., Calleri,A., Perma,G., Adorini,L., Luini,W., Mantovani,A., Van Damme,J., and Sozzani,S. (2003). Unique regulation of CCL18 production by maturing dendritic cells. J. Immunol. 170, 3843-3849.

Walz,A., Schmutz,P., Mueller,C., and Schnyder-Candrian,S. (1997). Regulation and function of the CXC chemokine ENA-78 in monocytes and its role in disease. J. Leukoc. Biol. 62, 604-611.

Wan,Y. and Bramson,J. (2001). Role of dendritic cell-derived cytokines in immune regulation. Curr. Pharm. Des. 7, 977-992.

Wang,W., Bacon,K.B., 01dham,E.R., and Schall,T.J. (1998). Molecular cloning and functional characterization of human MIP-1 delta, a new C-C chemokine related to mouse CCF-18 and CIO. J. Clin. Immunol. 18, 214-222.

Wang,Z.Q., Ovitt,C., Grigoriadis,A.E., Mohle-Steinlein,U., Ruther,U., and Wagner,E.F. (1992). Bone and haematopoietic defects in mice lacking c-fos. Nature 360, 741-745.

Wilkin,F., Duhant,X., Bruyns,C., Suarez-Huerta,N., Boeynaems,J.M., and Robaye,B. (2001). The P2Y11 receptor mediates the ATP-induced maturation of human monocyte-derived dendritic cells. J. Immunol. 166, 7172-7177.

Williams,M. and Jarvis,M.F. (2000). Purinergic and pyrimidinergic receptors as potential drug targets. Biochem. Pharmacol. 5 9 ,1173-1185.

Wilson,S., Bergsma,D.J., Chambers,J.K., Muir,A.I., Fantom,K.G., Ellis,C., Murdock,P.R., Herrity,N.C., and Stadel,J.M. (1998). Orphan G-protein-coupled receptors: the next generation of drug targets? Br. J. Pharmacol. 125, 1387-1392.

Wise,A., Gearing,K., and Rees,S. (2002). Target validation of G-protein coupled receptors. Drug Discov. Today 7, 235-246.

Wolpe,S.D., Davatelis,G., Sherry,B., Beutler,B., Hesse,D.G., Nguyen,H.T., Moldawer,L.L., Nathan,C.F., Lowry,S.F., and Cerami,A. (1988). Macrophages secrete a novel heparin-binding protein with inflammatory and neutrophil chemokinetic properties. J. Exp. Med. 167, 570-581.

Wu,L., Li,C.L., and Shortman,K. (1996). Thymic dendritic cell precursors: relationship to the T lymphocyte lineage and phenotype of the dendritic cell progeny. J. Exp. Med. 184, 903-911.

Wurtzen,P.A., Nissen,M.H., and Claesson,M.H. (2001). Maturation of dendritic cells by recombinant human CD40L-trimer leads to a homogeneous cell population with enhanced surface marker expression and increased cytokine production. Scand. J. Immunol. 53, 579-587.

Yang,D., Howard,O.M., Chen,Q., and Oppenheim,J.J. (1999).Immature dendritic cells generated from monocytes in the presence of TGF-beta 1 express functional C-C chemokine receptor 6. J. Immunol. 163,1737-1741.

Yasuda,H., Shima,N., Nakagawa,N., Yamaguchi,K., Kinosaki,M., Mochizuki,S., Tomoyasu,A., Yano,K., Goto,M., Murakami,A., Tsuda,E., Morinaga,T., Higashio,K., Udagawa,N., Takahashi,N.,

214 and Suda,T. (1998). Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to TRANCE/RANKL. Proc. Natl. Acad. Sci. U. S. A 95, 3597-3602.

Yoshida,R., Nagira,M., Imai,T., Baba,M., Takagi,S., Tabira,Y., AkagiJ., Nomiyama,H., and Yoshie,0. (1998a). EBIl-ligand chemokine (EEC) attracts a broad spectrum of lymphocytes: activated T cells strongly up-regulate CCR7 and efficiently migrate toward EEC. Int. Immunol. 10, 901-910.

Yoshida,R., Nagira,M., Kitaura,M., Imagawa,N., Imai,T., and Yoshie,0. (1998b). Secondary lymphoid-tissue chemokine is a functional ligand for the CC chemokine receptor CCR7. J. Biol. Chem. 275, 7118-7122.

Youn,B.S., Kim,S.H., Lyu,M.S., Kozak,C.A., Taub,D.D., and Kwon,B.S. (1997a). Molecular cloning and characterization of a cDNA, CHEMRl, encoding a chemokine receptor with a homology to the human C-C chemokine receptor, CCR-4. Blood 89, 4448-4460.

Youn,B.S., Zhang,S.M., Lee,E.K., Park,D.H., Broxmeyer,H.E., Murphy,P.M., Locati,M., Pease,J.E., Kim,K.K., Antol,K., and Kwon,B.S. (1997b). Molecular cloning of leukotactin-1: a novel human beta-chemokine, a chemoattractant for neutrophils, monocytes, and lymphocytes, and a potent agonist at CC chemokine receptors 1 and 3. J. Immunol. 159, 5201-5205.

Young,J.W. and Steinman,R.M. (1988). Accessory cell requirements for the mixed-leukocyte reaction and polyclonal mitogens, as studied with a new technique for enriching blood dendritic cells. Cell Immunol. Ill, 167-182.

Yu,X., Huang,Y., Collin-Osdoby,P., and Osdoby,P. (2003). Stromal cell-derived factor-1 (SDF-1) recruits osteoclast precursors by inducing chemotaxis, matrix metalloproteinase-9 (MMP-9) activity, and collagen transmigration. J. Bone Miner. Res. 18, 1404-1418.

Zhou,L.J. and Tedder,T.F. (1995). Human blood dendritic cells selectively express CD83, a member of the immunoglobulin superfamily. J. Immunol. 154, 3821-3835.

Zhu,K., Shen,Q., Ulrich,M., and Zheng,M. (2000). Human monocyte-derived dendritic cells expressing both chemotactic cytokines IL-8, MCP-1, RANTES and their receptors, and their selective migration to these chemokines. Chin. Med. J. (Engl. ) 775,1124-1128.

Zou,W., Borvak,L, Marches,F., Wei,S., Galanaud,P., Emilie,D., and Curiel,T.J. (2000). Macrophage-derived dendritic cells have strong Thl-polarizing potential mediated by beta- chemokines rather than IL-12. J. Immunol. 165, 4388-4396.

215

Reproduced with permission of copyright owner. Further reproduction prohibited without permission.