Novel Insights into Inflammatory Disturbed Bone Remodelling

Elin Kindstedt

Department of Odontology Umeå 2017

Umeå University Odontological Dissertations, New Series No: 138

Responsible publisher under Swedish law: the Dean of the Medical Faculty This work is protected by the Swedish Copyright Legislation (Act 1960:729) Dissertation for PhD ISBN: 978-91-7601-762-3 ISSN: 0345-7532 New series No: 138 Cover: Osteoclasts cultured on glass, stained to visualise filamentous actin Electronic version available at: http://umu.diva-portal.org/ Printed by: UmU Print Service, Umeå University Umeå, Sweden 2017

To my family

i

Table of Contents

Abstract .......................................................................................... iiiPopulärvetenskaplig sammanfattning på svenska ............................ vList of publications ......................................................................... viiAbbreviations ................................................................................ viiiIntroduction ..................................................................................... 1

Bone tissue ....................................................................................................................... 1Bone function, structure and composition ............................................................... 1Bone cells .................................................................................................................. 2Bone as a living tissue .............................................................................................. 5

Disturbed bone remodelling ........................................................................................... 7Definition and mechanisms ..................................................................................... 7Chemokines ............................................................................................................... 9

Osteopathological condititons ....................................................................................... 12Periodontitis ............................................................................................................ 12Rheumatoid arthritis (RA) ...................................................................................... 18Association between periodontitis and RA ............................................................ 19

Aims ............................................................................................... 20Methods .......................................................................................... 21

Humans .......................................................................................................................... 21Mice ............................................................................................................................... 22Cell isolation and culture .............................................................................................. 22Pam2-induced inflammation and bone resorption ..................................................... 24Gene expression analysis .............................................................................................. 24Protein expression analysis .......................................................................................... 25Histology and staining .................................................................................................. 26Osteoclast formation, migration and bone resorption ................................................ 26Flow cytometry ............................................................................................................. 27Radiographic assessment of bone loss ......................................................................... 27Statistical analysis ......................................................................................................... 28

Results and discussion ................................................................... 29Conclusions .................................................................................... 39Acknowledgements ........................................................................ 40References ..................................................................................... 43

ii

iii

Abstract

Bone is a dynamic tissue that is continuously remodelled, a process that requires equal amounts of osteoclastic bone resorption and osteoblastic bone formation. Inflammation may disturb the equilibrium and result in local and/or systemic bone loss. Negative bone mass balance occurs in several chronic inflammatory diseases, e.g. periodontitis and rheumatoid arthritis (RA). The aetiology of periodontitis is infectious, while RA is an autoimmune disease. Despite aetiological differences, an association between the two diseases has been established but it is not known if they are causally related. Periodontitis may develop when the inflammatory process, initially restricted to the gingiva (gingivitis), further invades the periodontium and causes bone resorption. The cellular and molecular mechanisms underlying the transition from gingivitis to periodontitis are not fully elucidated. Osteoclast formation is dependent on receptor activator of nuclear factor kappa B ligand (RANKL), but how osteoclast precursors are recruited to the jawbone is poorly understood. A family of cytokines named chemokines has been reported to possess such properties and increasing evidence points towards their involvement in the pathogenesis of chronic inflammatory diseases.

The overall aim of this thesis was to gain extended knowledge about the role of chemokines and a newly discovered family of leukocytes named innate lymphoid cells (ILCs) in periodontitis and concomitant inflammatory disturbed bone remodelling. Furthermore, the aim was also to study the association between periodontitis and RA.

We identified increased serum levels of monocyte chemoattractant protein (MCP)-1 and CCL11 in individuals with periodontitis. Moreover, a robust correlation between the two chemokines and periodontitis was detected in a weighted analysis of inflammatory markers, subject characteristics and periodontitis parameters. We detected higher MCP-1 levels in periodontitis tissue compared to non-inflamed. Furthermore we demonstrated that human gingival fibroblasts express MCP-1 and CCL11 in response to pro-inflammatory cytokines through NF-κB signalling. Using an inflammatory bone lesion model and primary cell cultures, we discovered that osteoblasts express CCL11 in vivo and in vitro and that the expression increased under inflammatory conditions. Osteoclasts did not express CCL11, but its high affinity receptor CCR3 was upregulated during osteoclast differentiation and found to co-localise with CCL11 on the surface of osteoclasts. Exogenous CCL11 was internalised in osteoclasts, stimulated the migration of osteoclast precursors and increased bone resorption in vitro.

iv

To analyse if periodontitis precedes RA we analysed marginal jawbone loss in dental radiographs taken in pre-symptomatic RA cases and matched controls. The prevalence of jawbone loss was higher among cases, and the amount of jawbone loss correlated with plasma levels of RANKL.

In the search of the newly discovered ILCs, we performed flow cytometry analyses on gingivitis and periodontitis tissue samples. We detected twice as many ILCs in periodontitis as in gingivitis. In addition we found RANKL expression on ILC1s (an ILC subset).

In conclusion, we demonstrated that CCL11 is systemically and locally increased in periodontitis and that the CCL11/CCR3 axis may be activated in inflammatory disturbed bone remodelling. We also found that marginal jawbone loss correlated with plasma levels of RANKL and preceded clinical onset of symptoms of RA. Furthermore, we demonstrated that ILCs are present in periodontitis and represent a previously unknown source of RANKL.

v

Populärvetenskaplig sammanfattning på svenska

Skelettet har flera viktiga funktioner i kroppen såsom att möjliggöra en upprätt hållning, utgöra fäste för muskler och mediera rörelse, skydda benmärgen och de inre organen samt reglera mängden av lösligt mineral i blodet. Med tiden uppstår mikroskador i skelettet vilket innebär att benvävnaden måste byggas om för att vara fortsatt funktionell. Ombyggnaden kallas remodellering och är en kontinuerlig process som huvudsakligen utförs av benbildande celler kallade osteoblaster och bennedbrytande celler kallade osteoklaster. Remodelleringen är strikt reglerad av olika signalmolekyler och under friska förhållanden råder jämvikt mellan mängden ben som bryts ner och mängden ben som bildas, vilket innebär att benmassan hålls konstant. Vid sjukdomar som medför långvariga inflammationsprocesser i benvävnad eller i närheten av benvävnad, exempelvis parodontit (tandlossningssjukdom) och ledssjukdomen reumatoid artrit (RA), kan den rådande jämvikten rubbas, vilket oftast resulterar i minskad benmängd. Vid parodontit är den bakomliggande orsaken till inflammationen bakterier som finns i placket på tänderna, men vid RA tros anledningen vara att immunförsvaret attackerar kroppsegna celler. Trots olikheterna delar de två sjukdomarna flera gemensamma drag med avseende på riskfaktorer, vilka signalmolekyler som återfinns i blodet samt hur inflammationsprocessen fortskrider. Parodontit föregås av gingivit (tandköttsinflammation). Hos vissa individer övergår gingivit till parodontit, en process som inkluderar nedbrytning av tandens stödjevävnader inklusive käkben. Det är inte helt klarlagt vilka celler och molekyler som finns närvarande vid gingivit respektive parodontit eller vilka mekanismer som ligger bakom skiftet mellan de två tillstånden. Det är sedan tidigare känt att molekylen RANKL är viktig för osteoklastbildning, men det är delvis okänt hur osteoklastförstadieceller rekryteras från blodcirkulationen till käkbenet. En grupp av molekyler kallade kemokiner, som även finns i förhöjda nivåer i blod vid parodontit och RA, har visat sig ha sådana egenskaper.

För att finna läkemedel som kan förhindra bennedbrytning till följd av den inflammatoriskt störda benremodellering som sker vid både parodontit och RA är det viktigt att studera sambandet mellan sjukdomarna och få en tydlig bild av vilka celler som är närvarande vid inflammationsprocessen. Det är även av betydelse att kartlägga vilka celler och molekyler som främjar rekrytering av osteoklastförstadieceller och bidrar till bennedbrytning.

Syftet med den här avhandlingen var att undersöka betydelsen av kemokiner vid inflammatoriskt störd benremodellering och vid parodontit samt att undersöka sambandet mellan parodontit och RA. För att skapa en tydligare bild av vilka cellertyper som är närvarande vid inflammationsprocessen vid parodontit

vi

undersöktes även förekomsten av en nyligen upptäckt celltyp vid namn ILC-immunceller (ILCs) samt om dessa celler uttrycker RANKL.

Först analyserades förekomsten av olika inflammatoriska signalmolekyler i blod från individer med parodontit samt från friska kontroller. Individer med parodontit hade förhöjda nivåer av kemokinerna MCP-1 och CCL11. Genom att använda en statistisk analysmetod som utöver inflammatoriska signalmolekyler även inkluderade kliniska variabler kunde ett samband mellan de två kemokinerna och parodontit påvisas. Vidare undersöktes möjliga ursprung till de i blodet förhöjda kemokinnivåerna genom att analysera tandkött från tänder med parodontit samt friskt tandkött. Vid parodontit uppmättes högre nivåer av MCP-1. Gingivala fibroblaster (en celltyp som producerar bindväv och ansvarar för tandköttets uppbyggnad) från människa bildade MCP-1 och CCL11 när de stimulerats med inflammationsfrämjande substanser, vilket krävde aktivering en intracellulär signaleringsväg kallad NF-κB. För att utreda betydelsen av CCL11 vid inflammatoriskt störd benremodellering analyserades bencellers bildning av CCL11 in vivo i skalltak från möss samt in vitro i cellodlingar. Osteoblaster bildade CCL11 in vivo och in vitro och bildningen ökade under inflammatoriska förhållanden. Osteoklaster bildade inte CCL11, men däremot fanns ett uttryck av receptorn CCR3, vilket är en mottagarmolekyl till CCL11. I vävnadssnitt från skalltak visades att CCL11 och CCR3 ser ut att binda till varandra på osteoklasternas yta. Dessutom hade CCL11 en positiv effekt på rekrytering av osteoklastförstadieceller och CCL11 som tillsattes till cellodlingar togs upp av osteoklaster och stimulerade benresorption.

För att studera sambandet mellan parodontit och RA analyserades käkbensförlust vid tänder med hjälp av röntgenbilder tagna på individer som senare utvecklade RA (pre-symptomatiska) samt matchade kontroller. De pre-symptomatiska individerna hade en högre grad av käkbenförlust och det fanns också ett samband mellan käkbensförlust och nivåer av RANKL i blodet.

Förekomsten av ILCs i tandkött från tänder med gingivit respektive parodontit analyserades med flödescytometri. Dubbelt så många ILCs återfanns vid parodontit än vid gingivit, varav majoriteten bestod av ILC1 (en undergrupp till ILCs). Vidare analyser visade att ILC1 cellerna bildar RANKL.

Sammanfattningsvis, vid parodontit finns förhöjda nivåer av CCL11 i vävnaden och i blodet, och interaktionen mellan CCL11 CCR3 kan vara av betydelse vid inflammatoriskt störd benremodellering. Käkbensförlust föregår RA och korrelerar med nivåer av den osteoklaststimulerande molekylen RANKL i blodet, vilket stödjer teorin om att det finns ett samband mellan de två sjukdomarna. De nyligen upptäckta ILCs återfinns vid både gingivit och parodontit och utgör dessutom en tidigare okänd källa till RANKL.

vii

List of publications

This thesis is based on the following papers, which will be referred to by their roman numerals (I-IV):

I. Boström, E. A*., Kindstedt, E*., Sulniute, R., Palmqvist, P., Majster, M., Holm, C. K., Zwicker, S., Clark, R., Önell, S., Johansson, I., Lerner, U. H. & Lundberg, P. 2015. Increased Eotaxin and MCP-1 Levels in Serum from Individuals with Periodontitis and in Human Gingival Fibroblasts Exposed to Pro-Inflammatory Cytokines. PLoS One, 10, e0134608.

II. Kindstedt, E., Holm, C. K., Sulniute, R., Martinez-Carrasco, I., Lundmark, R. & Lundberg, P. 2017. CCL11, a novel mediator of inflammatory bone resorption. Sci Rep, 7, 5334.

III. Kindstedt, E., Johansson, L., Palmqvist, P., Holm, C.K., Kokkonen, H., Rantapää Dahlqvist, S*. & Lundberg, P*. Marginal jawbone loss is associated with onset of rheumatoid arthritis and is related to plasma level of activator of nuclear factor kappa-B ligand (RANKL). Revised manuscript submitted to journal.

IV. Kindstedt, E., Holm, C.K., Palmqvist, P., Sjöström, M., Lejon, K*. & Lundberg, P*. Discovery of innate lymphoid cells in gingivitis and periodontitis. Manuscript.

*Contributed equally to the study

viii

Abbreviations

α-MEM alpha minimal essential medium ACPA anti-citrullinated protein antibodies ALP alkaline phosphatase ANOVA analysis of variance ATP adenosine triphosphate BCA bicinchoninic acid BMI body mass index BMM bone marrow macrophage BMP bone morphogenic protein BMU bone multicellular unit/bone metabolic unit BSP bone sialoprotein cAMP cyclic adenosine monophosphate CCP cyclic citrullinated peptide cDNA complementary deoxyribonucleic acid CEJ cemento-enamel junction c-Fms colony-stimulating factor 1 receptor CRP c-reactive protein DAB 3,3′-diaminobenzidine DAPI 4',6-diamidino-2-phenylindole D3 1,25(OH)2 vitamin D3 EDTA ethylenediaminetetraacetic acid ELAM-1 endothelial leukocyte adhesion molecule 1 ELISA enzyme-linked immunosorbant assay ERK extracellular signal-regulated protein kinase F-actin filamentous actin FBS foetal bovine serum GAPDH glyceraldehyde-3-phosphate dehydrogenase GATA3 GATA binding protein 3 G-protein guanine-nucleotide-binding protein HEPES hydroxyethyl piperazineethanesulfonic acid HLA human leukocyte antigen HRP horseradish peroxidase HSC hematopoietic stem cell ICAM-1 intracellular adhesion molecule 1 IFN interferon IGF insulin-like growth factor IL interleukin ILC innate lymphoid cell Lti lymphoid tissue inducer MCP-1 monocyte chemoattractant protein 1 M-CSF macrophage colony-stimulating factor MIP-1α macrophage inflammatory protein 1 alpha MMP matrix metalloproteinase mRNA messenger ribonucleic acid MSC mesenchymal stem cell NFATc1 nuclear factor of activated T cells, cytoplasmic 1 NF-κB nuclear factor kappa B OPG osteoprotegrin Pam2 pam2CSK4 PBMC peripheral blood monocytes

ix

PBS phosphate buffered saline PDL periodontal ligament PI3K phosphoinositide 3-kinase PLS partial least squares PTH parathyroid hormone RA rheumatoid arthritis RANK receptor activator of nuclear factor kappa B RANKL receptor activator of nuclear factor kappa B ligand RANTES regulated on activation, normal T cell expressed and secreted RF rheumatic factor ROC receiver operating characteristic RORyt retinoic acid receptor-related orphan receptor gamma t RPL13a ribosomal protein L13a RTqPCR quantitative real-time reverse transcription polymerase chain reaction Runx2 runt-related transcription factor 2 SDF-1 stromal cell-derived factor 1 S1P sphingosine-1-phosphate TGF transforming growth factor TNF tumour necrosis factor TRAP tartrate-resistant acid phosphatase TLR toll-like receptor

x

1

Introduction

Bone tissue

Bone function, structure and composition

Function From the small auditory ossicles to the great femur, the approximately 200 bones and associated cartilage forming our skeleton are the reinforcing bars of our body. Tendons originating from muscles attach to bone surfaces, and muscle contraction enables both movement and an upright posture. The skeleton surrounds and protects sensitive and vital inner organs, e.g. brain, heart, lungs, intestine and bone marrow. Furthermore, the skeleton can release or incorporate minerals in order to maintain mineral homeostasis in the body.

Structure Bone is a particularly specialised connective tissue that exists in two morphologically diverse forms: cortical (compact) bone and cancellous (spongy or trabecular) bone. The dense cortical bone constitutes an outer shell on most bones, surrounding a core of mesh-like cancellous bone (Figure 1), a structure that gives bone great mechanical strength and low weight in tandem. The external surface of the cortical bone (periosteum) contains blood vessels and nerves, whereas the inner surface (endosteum) faces the cancellous bone that harbours the bone marrow.

Composition Although cortical and cancellous bone tissue are structurally different from each other, both forms comprise a composition of an inorganic mineralised component (approximately 65%) and an organic matrix (approximately 35%). The inorganic mineralised component gives the skeleton its rigidity and strength, and consists almost exclusively of hydroxyapatite [Ca10(PO4)6 (OH2)], which is also found in tooth enamel and dentin. Type I collagen fibres are the major component of the organic matrix, constituting approximately 90%. The remaining 10% consists of proteoglycans (e.g. aggrecan, versican, decorin, biglycan), glycoproteins (e.g. alkaline phosphatase (ALP), osteonectin, tetranectin), glycoproteins with cell-attachment activities (e.g. bone sialoprotein (BSP), thrombospondins, fibronectin, vitronectin, osteopontin), γ-carboxy glutamic acid-containing proteins (e.g. matrix Gla protein, osteocalcin) and growth factors (e.g. transforming growth factor beta (TGFβ), bone morphogenetic proteins (BMPs) insulin-like growth factors (IGFs)). Although

2

these non-collagenous proteins are not as abundant as collagenous protein, they all have unique and important effector functions (Clarke, 2008).

Figure 1. Bone structure, macroscopic (left) and microscopic (right) view.

Bone cells Apart from the inorganic mineralised component and the organic matrix, bone tissue contains four cell types that are responsible for its formation, maintenance, reconstruction, reshaping and degradation. Osteoblasts, bone lining cells and osteoclasts are present on bone surfaces, whereas osteocytes are located within the mineralised bone matrix (Figure 2).

Osteoblasts Responsible for bone formation is the osteoblast, a cell type derived from mesenchymal stem cells (MSCs). Certain transcription factors require activation in order for MSCs to differentiate within the osteoblastic lineage and the very same transcription factors simultaneously inhibit differentiation of MSCs into cell types of the adipocytic and chondrocytic lineages (Aubin and Heersche, 2000). Runt-related transcription factor 2 (Runx2) is crucial for osteoblastogenesis (Komori et al., 1997, Otto et al., 1997), but not sufficient to

3

fully support the differentiation. Osterix, which is a downstream target for Runx2, also requires activation (Nakashima et al., 2002).

At a functional level, the mature osteoblast is characterised by its ability to synthesise bone tissue, including the production of organic matrix and subsequent mineralisation. The formation of organic matrix, which is also referred as osteoid, prevails mineralisation. A thin layer of osteoid covers all non-resorbing bone surfaces. Osteoblasts are phenotypically recognised by their ability to form bone matrix proteins (e.g. collagen and osteocalcin) and their strong expression of alkaline phosphatase (ALP), an enzyme associated with bone mineralisation (Lian and Stein, 1995). Although osteoblasts are mainly responsible for bone formation, they are also involved in bone resorption processes (Rodan and Martin, 1982). Osteoblasts are provided with receptors for the calcium regulating hormones parathyroid hormone (PTH) and 1,25(OH)2 vitamin D3 (D3) and possess the ability to express and release macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor kappa B ligand (RANKL), cytokines that are crucial for osteoclastogenesis (Horwood et al., 1998).

Osteocytes The deposition of bone matrix by osteoblasts is usually polarised towards the bone surface but sometimes the deposition becomes pericellular and as a result, osteoblasts are trapped in the bone matrix and thence referred to as osteocytes. Osteocytes are stellate shaped and reside with their cell bodies in lacunae, from which long cytoplasmic processes extend to make contact with neighbouring osteoblasts and osteocytes. Communication between osteocytes within the network can either be direct through gap junctions or indirect through paracrine signalling. Osteocytes are able to sense mechanical and chemical signals from the environment and regulate both bone formation and bone resorption through communication with osteoblasts and osteoclasts (Schaffler et al., 2014). Furthermore, osteocytes express receptors for PTH and D3 and are suggested being a substantial source of RANKL (Xiong et al., 2015).

Bone lining cells Bone lining cells are flat and elongated cells that cover inactive bone surfaces, i.e. surfaces not subjected to bone remodelling, and they are particularly evident in the adult skeleton. In terms of matrix production they are considered to be inactive, and the function of these cells is poorly understood. It is suggested that bone lining cells may be involved in the propagation of the activation signal that initiates bone resorption (Miller et al., 1989). It is also suggested that bone lining cells form a canopy above the resorption pit and compose a bridge between the events of bone resorption and bone formation (Delaisse, 2014).

4

Osteoclasts The osteoclast is a multinucleated and highly motile cell that possesses the unique ability to resorb mineralised bone. Unlike the other bone cells, osteoclasts are derived from hematopoietic stem cells (HSCs) of the monocyte-macrophage lineage. The process of osteoclastogenesis is complex and it has been established that a direct cell-to-cell contact between osteoblasts/stromal cells and osteoclast precursors involving ligand-receptor interaction is indispensable for the formation of osteoclasts (Suda et al., 1999).

Osteoblasts and stromal cells are able to produce M-CSF, a cytokine required for proliferation of osteoclast precursors and their further differentiation and fusion into osteoclasts (Felix et al., 1994, Tanaka et al., 1993). Accordingly, mice with an additional insertion of thymidine in the coding region of the M-CSF gene are unable to produce M-CSF and hence lack functional osteoclasts and develop osteopetrosis (Yoshida et al., 1990). The binding of M-CSF to its receptor colony-stimulating factor 1 receptor (c-Fms) results in downstream intracellular signalling through phosphoinositide 3-kinase (PI3K)-Akt and Grb2- extracellular signal-regulated protein kinase (ERK), resulting in activation of transcription factor PU.1, which in turn regulates genes associated with survival, proliferation and differentiation of osteoclast precursors (Souza and Lerner, 2013).

Further differentiation and terminal fusion of preosteoclasts primed by M-CSF into active osteoclasts are mediated by the interaction of RANKL and its cognate receptor RANK (Boyce, 2013). Osteoblasts, osteocytes, stromal cells, T-lymphocytes, synovial fibroblasts and periodontal ligament (PDL) cells express RANKL, a protein belonging to the tumour necrosis factor (TNF) superfamily that can be either membrane-bound or soluble. The interaction between RANKL and RANK activates the transcription factor nuclear factor kappa-B (NF-κB), which in turn induces the expression of nuclear factor of activated T-cells cytoplasmic 1 (NFATc1), ultimately enhancing the expression of several genes important for commitment to the osteoclastic cell phenotype, fusion of preosteoclasts and activation of mature osteoclasts (Takayanagi et al., 2002). Activation of additional co-stimulatory pathways including immunoglobulin-like receptors is needed to fully support osteoclast differentiation. Mature osteoclasts are multinucleated and express the osteoclast markers tartrate resistant acid phosphatase (TRAP) and Cathepsin K (Boyle et al., 2003). Deletion of either RANKL or RANK in mice results in complete absence of osteoclasts and subsequent osteopetrosis, although the number of macrophages remains unaffected (Dougall et al., 1999, Kong et al., 1999). Osteoprotegrin (OPG) is a soluble protein serving as a decoy receptor for RANKL that can block the interaction between RANKL and RANK. Mice overexpressing OPG display increased bone mass and reduced number of osteoclasts (Simonet et al., 1997),

5

which indicate that OPG is an important suppressing regulator of osteoclastogenesis. Additional molecules including hormones, cytokines, interferons and chemokines are thought to regulate the ratio between RANKL and OPG, which in turn regulates osteoclast formation, activation and resorption (Boyce, 2013).

Figure 2. Schematic illustration of bone cells and their origin.

Bone as a living tissue Throughout an individual’s lifetime, the skeleton is constantly altered. After the initial ossification of the embryonic skeleton, bone undergoes growth, modelling and remodelling. Growth refers to the longitudinal and radial expansion of the bones that occurs during the first two decades of life whereas modelling represents the process by which bones gradually adjust their shape and mass in order to meet physiological and mechanical requirements. During bone modelling, the processes of bone formation and bone resorption are not necessarily coupled. Bone modelling occurs more frequently in the skeleton of children and adolescents than in adults. Bone remodelling is a coupled process by which bone tissue is rebuilt to maintain bone strength and mineral homeostasis in the body. The remodelling process relies on a fine balance between bone resorbing osteoclasts and bone forming osteoblasts. The skeleton is remodelled from birth to death and is affected by hormonal changes, diseases and aging (Clarke, 2008).

6

The remodelling cycle Bone remodelling is confined to a restricted area of the bone surface and is mediated by an anatomic structure of bone cells named the basic multicellular unit/bone metabolic unit (BMU). The lifetime of a BMU is substantially longer than its individual cells and consequently, new cells have to be recruited to the BMU continuously (Jilka, 2003). After the BMU is created it begins to resorb old bone and subsequently new bone is formed. Short after, the BMU dissolves and leaves newly formed bone covered by a layer of inactive bone lining cells.

Initiation: It is not known how a BMU recognises a prospective remodelling site. It is suggested that remodelling sites may develop randomly but it can also be targeted to areas that require repair (Burr, 2002, Parfitt, 2002). The remodelling cycle begins with the degradation of the non-mineralised osteoid that covers all mineralised bone, a process that is mediated by proteolytic enzymes (e.g. matrix metalloproteinases (MMPs)) released by activated osteoblasts and bone lining cells. After removal of the osteoid, the bone lining cells withdraw in order to let the bone resorbing osteoclasts adhere to the mineralised surface (Ott, 2002). The activated osteoblasts also express and release molecules with chemoattractant properties that guide mononuclear osteoclast precursors to bone surfaces where they further differentiate and become active osteoclasts due to M-CSF and RANKL.

Resorption: Upon arrival to the resorption site, the cytoskeletons of the osteoclasts undergo rearrangements and the osteoclasts attach to the bone surface using a specialised part of the cell membrane named “the sealing zone”. Adhesion is mediated through binding between extracellular matrix receptors expressed on osteoclasts and extracellular matrix proteins embedded in bone (e.g. collagen, vitronectin, osteopontin, BSP) (Nesbitt et al., 1993). The sealing zone defines an area underneath the osteoclast where resorption occurs, known as Howship’s lacuna. The plasma membrane facing the lacuna is considered to be the actual resorbing organ of the osteoclast and is organised as a ruffled border formed by acidic intracellular vesicles. The bone resorption process begins with acidification of the lacuna due to hydrogen ions delivered by V-type adenosine triphosphate (ATP)-ase pumps and secretion of acidic vesicles that fuse with the cell membrane in the ruffled border. The acidic environment in the lacuna causes dissolution of bone tissue mineral. Either overlapping, or after this event, proteolytic enzymes (MMPs, Cathepsin K and TRAP) are released into the lacuna in order to degrade the organic matrix (Vaananen and Zhao, 2002).

Formation and mineralisation: After completion of the resorption phase the osteoclast most likely undergoes apoptosis, although there are studies indicating that an osteoclast can participate in more than one resorption cycle (Kanehisa

7

and Heersche, 1988). Unknown local signalling processes named “coupling processes” generate an osteogenic environment at the remodelling sites and attract osteoblasts to the resorption pit. It is proposed that these coupling factors, i.e. TGFβ, may be embedded in the bone matrix and released upon resorption or released by the osteoclast itself (Bonewald and Mundy, 1990, Hock et al., 1988, Locklin et al., 1999, Martin and Sims, 2005).

Activated osteoblasts invade the resorption pit and begin to produce type I collagen and non-collagenous proteins forming the osteoid. Ultimately, the osteoblasts mediate mineralisation of the osteoid in a manner that is currently not fully understood. It is hypothesised that osteoblasts transport small membrane-bound matrix vesicles containing ALP into the collagen network. The vesicles mediate concentration of calcium and phosphatase leading to the formation of crystals within the vesicles. The crystals are subsequently transferred outside the osteoblast and occupy available space between the collagen fibres. In addition, ALP enzymatically destroy mineralisation inhibitors such as pyrophosphate or proteoglycans, resulting in free organic phosphatase that can react with calcium to form hydroxyapatite (Anderson, 2003).

Disturbed bone remodelling

Definition and mechanisms

Definition Normal bone remodelling, which prevails in health, requires equal amounts of osteoclastic bone resorption and osteoblastic bone formation in order to preserve bone mass. Several pathological conditions, in particular longstanding chronic inflammatory diseases, can disrupt the fine balance and manifest as alterations in bone mass both locally and systemically (Agrawal et al., 2011). Inflammatory processes adjacent to bone or in bone itself often result in reduced bone mass, as in the case of periodontitis, rheumatoid arthritis (RA), osteoarthritis and the majority of tumours metastasising to bone. However, inflammatory processes in some cases result in increased bone mass, a phenomenon that is denoted sclerosis and can be observed in apical periodontitis and some bone metastasising tumours. Both bone formation and bone resorption processes are increased regardless of whether the result is increased or reduced bone mass, as it is the relative ratio between bone formation and bone resorption that determines the outcome. There is no common nomenclature to describe disturbances in bone remodelling due to inflammation and will in this thesis be referred to as inflammatory disturbed bone remodelling.

8

Mechanisms At a molecular level, the interplay between bone formation and bone resorption is mainly regulated by the RANKL/RANK/OPG ratio, which in turn is affected by several factors including cytokines, hormones, growth factors, prostaglandins, neuropeptides and skeletal loading. The listed factors may also exert direct effects on bone cells. The effects of cytokines on osteoclast formation are diverse. It has been demonstrated that TNF-α stimulates RANKL expression by osteoblasts (Boyce et al., 2005), and binding to its receptor results in activation of Nfatc1, thus directly promoting osteoclast formation in vitro (Kobayashi et al., 2000) but not in vivo (Li et al., 2000). Interleukin (IL)-1β was the first cytokine described to promote osteoclast formation (Dewhirst et al., 1985), and similar to TNF-α, is stimulates bone resorption in vitro (Gowen et al., 1983) and upregulates RANKL in stromal cells and osteoblasts (Hofbauer et al., 1999). Both TNF-α and IL-1β stimulate the expression of IL-6 by osteoblasts, which in turn induces osteoclast formation and bone resorption (Ishimi et al., 1990). Other cytokines that have been described promote osteoclast formation and bone resorption are IL-8, IL-11, IL-17, IL-20, IL-23, IL-32 and IL-34 whereas IL-4, IL-7, IL-10, IL-12, IL-13, IL-18, IL-27 and IL-33 are suggested to have inhibitory effects (Souza and Lerner, 2013). Among hormones, PTH and D3 are known mediators of osteoclast formation through upregulation of RANKL in osteoblasts (Huang et al., 2004, Takahashi et al., 2014). On the contrary, oestrogen inhibits bone resorption by suppressing the expression of TNF-α, IL-1β and IL-6. Moreover, oestrogen also induces the expression of OPG in osteoblasts and stromal cells (Saika et al., 2001). Growth factors are crucial for proliferation and differentiation of osteoblasts and are critically involved in bone formation and skeletal maintenance in adults (Linkhart et al., 1996). Prostaglandins are derived from arachidonic acid and are known mediators of bone resorption (Dietrich et al., 1975). The nervous system also participates in the regulating of bone cell functions (Lerner and Persson, 2008, Lundberg et al., 1999). The fact that immobility results in reduced bone mass suggest that mechanical loading is crucial for normal bone remodelling. It is suggested that osteocytes incorporated in the skeleton can detect mechanical loading and transmit the signal to other bone cells (Tatsumi et al., 2007). The complexity of this regulatory system is striking and it is suggested that an imbalance in the above-described factors may favour RANKL expression and bone resorption, ultimately resulting in bone loss.

Preceding osteoclast formation and subsequent bone resorption is the recruitment of leukocytes to the inflammation site. Leukocytes are primarily responsible for combating infection/injury, but they are also involved in bone metabolism since they are able to act as regulators of both osteoblast and osteoclast activity. Moreover, monocytes/macrophages represent a type of leukocyte with the ability to differentiate into osteoclasts. Inflammatory

9

disturbed bone remodelling with predominant bone resorption, which is seen in the majority of osteopathological conditions, is due to increased formation and activity of mature osteoclasts. The mechanisms by which osteoclast precursors of the monocyte/macrophage lineage are guided from the bone marrow to remodelling sites/inflammation sites on bone surfaces, although not fully understood, involves locally produced chemokines (Galliera et al., 2008).

Chemokines

Nomenclature, structure and function Chemokines represent a family of closely related cytokines with the ability to guide leukocytes out of the blood stream and to the inflammation site by chemotaxis, a process named migration or homing. Depending on the molecular structure, chemokines are divided into four different subfamilies: XC, CXC, CX3C and CC. Chemokine nomenclature is based on family, followed by an “L” (for “ligand”) and an individual number. However, many chemokines also retain an older name that denotes its function, for example monocyte chemoattractant protein-1 (MCP-1) or CCL2. A total number of approximately 50 chemokines and about 15 chemokine receptors have been reported (see Table 1). Thus, chemokine-chemokine receptor interactions are rarely specific. In most cases a single chemokine may bind to several receptors, whereas a single chemokine receptor can transduce signals for several cytokines. These features of chemokines and their cognate receptors enable high-precision recruitment of leukocytes to inflammatory processes (Zlotnik and Yoshie, 2000).

Chemokines exert their biological functions by interacting with G protein-linked trans-membrane receptors expressed on the surface of the target cells. The receptor is linked to the internal cytoskeleton of the cell and receptor activation results in cytoskeletal rearrangements and increased mobility towards the chemokine gradient. Based on functional characteristics, two different types of chemokines can be distinguished: basal and inflammatory. Basal chemokines are constitutively produced, recruited to, and expressed in lymphoid tissues where they are involved in maintaining tissue homeostasis. On the contrary, inflammatory chemokines are locally produced in response to inflammatory reactions. Disturbances in chemokine expression/function within tissue compartments might lead to a prolonged inflammatory response, which in turn promotes development of chronic inflammation (Mölne and Wold, 2007). As previously described, recruiting leukocytes is the main task of chemokines, but increasing evidence point towards the involvement of chemokines in other processes (e.g. fibrosis, tissue remodelling and angiogenesis) and conditions (vascular diseases, neoplasias, allergy, transplant rejection and auto-immunity) (Yadav et al., 2010).

10

Systematic name

Human synonym Mouse synonym Main receptor

XC Family XCL1 Lymphotactin/SCM-

1α/ATAC Lymphotactin XCR1

XCL2 SCM-1β Unknown XCR1 CXC Family CXCL1 GROα/MGSA-α GRO/KC? CXCR2 CXCL2 GROβ/MGSA-β GRO/KC? CXCR2 CXCL3 GROγ/MGSA-γ GRO/KC? CXCR2 CXCL4 PF4 PF4 Unknown CXCL5 ENA-78 LIX? CXCR2 CXCL6 GCP-2 CKα-3 CXCR1, CXCR2 CXCL7 NAP-2 Unknown CXCR2 CXCL8 IL-8 Unknown CXCR1, CXCR2 CXCL9 MIG MIG CXCR3 CXCL10 IP-10 IP-10 CXCR3 CXCL11 I-TAC Unknown CXCR3 CXCL12 SDF-1α/β SDF-1 CXCR4 CXCL13 BLC/BCA-1 BLC/BCA-1 CXCR5 CXCL14 BRAK/bolekine BRAK Unknown CXCL15 Unknown Lungkine Unknown CXCL16 SR-PSOX SR-PSOX CXCR6 CX3C Family CX3CL1 Fractalkine Neurotactin CX3CR1 CC Family CCL1 I-309 TCA-3, P500 CCR8 CCL2 MCP-1/MCAF MCP-1, JE? CCR2 CCL3 MIP-1α/LD78α MIP-1α CCR1, CCR5 CCL4 MIP-1β MIP-1β CCR5 CCL5 RANTES RANTES CCR1, CCR3,

CCR5 (CCL6) Unknown C10, MRP-1 Unknown CCL7 MCP-3 MARC? CCR1, CCR2,

CCR3 CCL8 MCP-2 MCP-2 CCR3 CCL9/10 Unknown MRP-2, CCF18, MIP-1Υ Unknown CCL11 eotaxin-1 eotaxin-1 CCR2, CCR3,

CCR5 (CCL12) Unknown MCP-5 CCR2 CCL13 MCP-4 Unknown CCR2, CCR3 CCL14 HCC-1 Unknown CCR1 CCL15 HCC-2/Lkn-1/MIP-1δ Unknown CCR1, CCR3 CCL16 HCC-4/LEC LCC-1 CCR1 CCL17 TARC TARC CCR4 CCL18 DC-CK1/PARC/AMAC-1 Unknown CCR3 CCL19 MIP-3β/ELC/exodus-3 MIP-3β/ELC/exodus-3 CCR7 CCL20 MIP-3α/LARC/exodus-1 MIP-3α/LARC/exodus-1 CCR6 CCL21 6Ckine/SLC/exodus-2 26Ckine/SLC/exodus-

2/TCA-4 CCR7

CCL22 MDC/STCP-1 ABCD-1 CCR4 CCL23 MPIF-1/ Ckβ-8 Unknown CCR1 CCL24 MPIF-2/eotaxin-2 Unknown CCR3 CCL25 TECK TECK CCR9 CCL26 Eotaxin-3 Unknown CCR10 CCL27 CTACK/ILC ALP/CTACK/ILCESkine CCR10 CCL28 MEC MEC/CCK1/Scya28 CCR3, CCR10 Table 1. Chemokines and chemokine receptors

11

Recruitment of osteoclast precursors The interaction between chemokines and their cognate receptors is important for the migration of osteoclast precursors to bone surfaces. Some of the more extensively researched chemokines are described below.

Stromal derived factor-1 (SDF-1) or CXCL12 is a chemokine expressed by osteoblasts, stromal cells and bone endothelial cells (Ponomaryov et al., 2000) and its cognate receptor CXCR4 is expressed on osteoclast precursors and on osteoclasts. SDF-1 has been demonstrated to promote chemotactic recruitment of osteoclast precursors and their further differentiation toward and osteoclastic phenotype, ultimately increasing osteoclast formation and bone resorption (Wright et al., 2005).

Monocyte chemoattractant protein-1 (MCP-1) or CCL2 has a documented effect on recruitment of monocytes (Galliera et al., 2008). Moreover, in an in vivo mouse model, osteoblasts constitute the main source of MCP-1 in inflamed jawbone and the number of monocytes/macrophages correlates with the number of MCP-1 positive cells (Rahimi et al., 1995). MCP-1 interacts with its receptor CCR2, which is expressed by osteoclasts and in RANKL-induced osteoclast differentiation (Kim et al., 2006). MCP-1/CCR2 interaction promotes fusion of preosteoclasts (Kim et al., 2005, Li et al., 2007). In addition, the absence of either MCP-1 or CCR2 results in increased bone mass (Sul et al., 2012, Binder et al., 2009).

Macrophage inflammatory protein 1-α (MIP-1α), systematically named CCL3, is able to utilise either CCR1 or CCR5 for its effects on osteoclast formation. The chemotactic effect of MIP-1α on osteoclast precursors has been established (Votta et al., 2000). Accordingly, neutralising antibodies against CCR1 or CCR5 inhibits MIP-1α-induced osteoclast formation (Oba et al., 2005). The CCL3 receptor CCR1 is also important for the differentiation of osteoblasts, and CCR1-deficiancy results in osteopenia (Hoshino et al., 2010).

Interleukin (IL)-8 or CXCL8 is an interleukin that upregulates RANKL in mouse osteoblastic cells and is suggested to cause osteoclast formation in vitro (Bendre et al., 2003). Moreover, blocking IL-8 with an antibody efficiently inhibits osteoclast formation induced by conditioned medium from breast cancer cells added to peripheral blood monocytes (Bendre et al., 2005).

Regulated on Activation Normal T cell Expressed and Secreted (RANTES) or CCL5 is a chemokine produced by a variety of cell types including osteoclasts. It is suggested that osteoclastic RANTES promote survival and chemotaxis of osteoblasts through interaction with its cognate receptor CCR5 (Yano et al., 2005). On the contrary, RANTES is also reported to stimulate chemotaxis of

12

preosteoclasts (Votta et al., 2000). Mice lacking functional RANTES display osteopenia associated with decreased bone formation and increased osteoclastogenesis. In addition, trabecular bone surfaces lack coverage of bone lining cells and osteoblasts (Wintges et al., 2013), which indicate that the main effects of RANTES are osteogenic (Liu et al., 2014).

It is also suggested that sphingosine-1-phosphate (S1P), not a chemokine but a lipid mediator capable of inducing chemotaxis, mediates the migration of osteoclast precursors from the bone marrow to the circulation (Ishii et al., 2009). Injection of a S1P agonist suppresses arthritis and osteoporosis in a mouse model, likely by facilitating osteoclast recirculation into blood and reducing the number of mature osteoclasts attached to the bone surfaces (Kikuta et al., 2011).

Osteopathological condititons

Periodontitis

Anatomy of the periodontium The tooth supporting tissues, collectively referred to as the periodontium, represent a functional unit that consist of four principal components: the gingiva, the periodontal ligament, the root cementum and the alveolar bone (Figure 3). The main function of the periodontium is to attach the tooth to the jaws, but it also offers nutrition to the tooth and serves as a barrier against bacterial challenges provided by the oral microbiota. Moreover, sensory receptors in the periodontium play an essential role in the regulation of masticatory function via signalling through the trigeminal nerve.

Consisting of epithelium and underlying connective tissue, the gingiva is the part of the masticatory mucosa that covers the alveolar bone and surrounds the teeth. The most coronal part of the gingiva is named the free gingiva and it extends from the gingival margin to a level corresponding to the cemento-enamel junction (CEJ) of the tooth. The gingiva apical to the free gingiva is named the attached gingiva, and it is continuous with the alveolar mucosa. The gingiva attains its scalloped shape in conjunction with tooth eruption and the shape adapts to changes in the marginal bone.

Three different types of gingival epithelium can be identified based on origin, location and composition: oral epithelium, sulcular epithelium and junctional epithelium. The oral epithelium is a stratified, squamous keratinising epithelium that faces the oral cavity and extends from the gingival margin to the alveolar mucosa. The sulcular epithelium is structurally similar to the oral

13

epithelium and faces the tooth. It is in close contact, but not attached to the tooth surface, and reaches from the gingival margin to the most coronal part of the junctional epithelium. The junctional epithelium is tightly attached to the tooth surface through hemidesmosomes and extends approximately two millimeters below the most apical part of the sulcular epithelium. In order to allow diffusion of leukocytes into the gingival crevice, the junctional epithelium is only a few cell layers thick (Bartold et al., 2000). The gingival connective tissue constitutes the vast majority of the gingiva. The major component of the connective tissue and also responsible for its structure and stability is a network of collagen fibres. In addition to collagen fibres, small amounts of proteoglycans and glycoproteins are also present in the connective tissue. The extracellular matrix is produced by gingival fibroblasts, a cell type of mesenchymal origin that is the predominant cell type in the connective tissue. Other cell types encountered in the gingiva are endothelial cells, mast cells, macrophages, polymorphonuclear leukocytes (PMNs), lymphocytes and plasma cells. The gingival connective tissue attaches to the alveolar bone and the root cementum by bundles of collagen fibres and hence also provides attachment for the tooth (Bartold et al., 2000).

The periodontal ligament represent a connective tissue situated in the space between the root surface and the alveolar bone. Bundles of collagen fibres extend from the root cementum and are incorporated in the alveolar socket wall. These fibres not only provide tooth attachment and a certain degree of mobility, they also distribute masticatory forces and act as shock absorbers. Cells present in the periodontal ligament are PDL cells, osteoblasts, osteoclasts and epithelial cells, among others (Lindhe et al., 2008). The periodontal ligament cell is a fibroblast-like cell with some osteoblastic features (Jonsson et al., 2011).

The root cementum covers the root surfaces. It is a mineralised tissue that resembles bone in its content of hydroxyapatite and organic matrix including collagen fibres. Different from bone, cementum is absent of both blood vessels and innervation and it does not undergo resorption or remodelling. The coronal and middle part of the root is covered by acellular cementum, in which collagen fibres extending from the alveolar bone are incorporated. In addition to fibres, the cellular cementum of the apical third of the root also contains cementoblasts, cells that continue to form cementum as long as the tooth is functional (Lindhe et al., 2008).

The alveolar bone comprises the parts of the maxilla and mandible that form the alveolar socket, the bone tissue that surrounds and supports the teeth. The alveolar socket is lined with cortical bone in which bundles of periodontal ligament fibres are attached. The spaces between the sockets and the cortical

14

jawbone are occupied with cancellous bone. Alveolar bone displays high remodelling rate due to forces elicited by mastication (Sodek and McKee, 2000).

Figure 3. Schematic illustration of the periodontium and its components.

Pathogenesis of periodontitis Periodontitis is a chronic inflammatory disease characterised by inflammation in the gingiva with concomitant loss of jawbone. The severe forms of periodontitis affect approximately 10% of the adult population and thus, the global burden of the disease is comprehensive (Kassebaum et al., 2014). Efforts have been directed towards elucidating the pathogenesis of periodontitis during the later half of the 20th century, but parts of it still remains elusive. It is established that the disease is initiated when microorganisms in the dental biofilm break the gingival barrier, resulting in an antigenic challenge that exceeds the resistance of the host’s immune response. The microbial invasion includes release of endotoxins (e.g. lipopolysaccharides) and enzymes (e.g. proteinases and leukotoxins), which either directly or indirectly harm the

15

periodontium (Kinane et al., 2008). Indirect effects require activation of the host’s innate and adaptive immune response with subsequent inflammation.

In a first step, microbial components activate and stimulate tissue resident cells (e.g. mast cells, macrophages and gingival fibroblasts) to release TNF-α and vasoactive amines that increase vascular permeability and the expression of adhesion molecules on endothelial cells. The adhesion molecules (including intercellular adhesion molecule-1 (ICAM-1), endothelial leukocyte adhesion molecule-1 (ELAM-1) and p-selectin) participate in recruiting leukocytes to the inflamed periodontal tissues, a process called homing. PMNs dominate the infiltration in the early phases of periodontal inflammation and they release lysosomal enzymes, which contribute to both connective tissue degradation and microbial killing. As connective tissue is degraded, it looses its entity, which allows further infiltration of lymphocytes and macrophages. Up to this point, when the periodontal inflammation is confined within the gingiva, the condition is named gingivitis. Gingivitis will in some disease-susceptible individuals further develop into periodontitis, which is characterised by irreversible breakdown of tooth supporting tissues including alveolar bone. Throughout the inflammatory process, the resident cells maintain their production of pro-inflammatory cytokines (e.g. TNF-α, IL-1β, IL-8, IL-6 and prostaglandins), which in turn stimulate fibroblasts to produce connective tissue degrading matrix metalloproteinases (MMPs). As connective tissue is degraded, the sulcular epithelium proliferates apically along the root and replaces the junctional epithelium, thus creating a deepened periodontal pocket. Furthermore, the pro-inflammatory cytokines elicit a cascade of osteotropic factors from fibroblasts, osteoblasts, stromal cells and leukocytes, ultimately leading to accumulation of osteoclast precursors and activation of osteoclasts through upregulation of RANKL (Yucel-Lindberg and Bage, 2013) (Figure 4).

The sequence of events described above is affected by environmental factors (e.g. smoking), diseases (e.g. diabetes mellitus) and hormonal changes (e.g. pregnancy) (Palmer and Soory, 2008). The presence of bacteria is undoubtedly necessary, but insufficient to alone cause disease, which indicate that the host’s susceptibility is decisive for disease development. An individual susceptibility to the disease has long been accepted and it is currently not known why some individuals do not develop periodontitis despite microbial load and presence of gingivitis for extended time periods while others rapidly loose tooth supporting alveolar bone (Loe et al., 1986, Page et al., 1997). It is suggested 50% of the risk of developing periodontitis is attributable to gene polymorphisms (Michalowicz et al., 1991, Hassell and Harris, 1995). Furthermore, the mechanisms underlying the transition from gingivitis to periodontitis and the cellular composition of both conditions have not yet been elucidated. It is known that the extent of the inflammatory infiltrate and the amount of lymphocytes increases with the

16

severity of the periodontal inflammation. In a meta-analysis performed by Berglundh et al., it was concluded that in the established periodontitis lesion, B cells represent the predominant cell type. However, cells of unknown cell type represent a considerable amount of 12 % (Berglundh et al., 2011).

Figure 4. Schematic illustration of the current concept of the aetiology and pathogenesis of periodontitis. Adapted from Page and Kornman 1997.

Innate lymphoid cells A newly discovered group of cells named innate lymphoid cells (ILCs) has gained increased attention within the field of immunology in the last decade. These cells are morphologically similar to lymphocytes and mirror the phenotypes and certain functions of T cells, but in contrast, they lack antigen-specific receptors and do not undergo clonal expansion. HSCs are able to differentiate upon activation of particular transcription factors to form three different ILC subsets, each characterised by a distinct cytokine expression pattern (Mjosberg and Spits, 2016). Briefly, group 1 ILCs (ILC1), including the conventional cytotoxic natural killer (NK) cell, depend on T-bet and produce interferon (IFN)-γ. Group 2 ILCs (ILC2) depend on GATA binding protein 3 (GATA3) and produce type 2 cytokines e.g. IL-5 and IL-13. Group 3 ILCs (ILC3) depend on RAR-related orphan receptor (ROR)-γt and can be further divided into lymphoid tissue-inducer (LTi) cells and NCR+ ILC3 or NCR- ILC3 depending on the expression of NKp44. ILC3s are associated with production of

17

IL-22 (NCR+) and IL17 (NCR- and NCR+) in response to stimulation by IL-1β and IL-23 (Konya and Mjosberg, 2015) (Figure 5). The absence of antigen-specific receptors makes ILCs incapable of directly recognising pathogens. Instead, ILCs respond to epithelium-derived signals (e.g. IL-25, IL-33) that are released from infected or damaged mucosa or to molecules released by myeloid cells (e.g. IL-1β, IL-12, IL-23) with the production of an array of effector cytokines that shape the subsequent immune responses (Eberl et al., 2015, Thaiss et al., 2016). The overall function of ILCs, although not fully elucidated, is to mediate tissue homeostasis and pathology at mucosal surfaces throughout the body. It is suggested that ILCs can be either harmful or protective depending on the context, and there is a notable shift in ILC numbers when comparing health and disease (Hazenberg and Spits, 2014). ILC1s are accumulated in, and possibly contribute to the pathogenesis of, chronic gut inflammation, more particularly Chron’s disease (Bernink et al., 2013). ILC2s participate in type-2-mediated allergy of the respiratory system and skin (Konya and Mjosberg, 2016). In addition, ILC2s produce amphiregulin, a protein important for tissue regeneration (Monticelli et al., 2011). LTi cells are essential for embryonic and postnatal development of lymphoid organs (Bar-Ephraim and Mebius, 2016). ILC3s are distributed in the intestine where they can either promote tissue homeostasis through IL-22 or mediate gut inflammation through IL-17 (Sonnenberg et al., 2011, Sonnenberg, 2014). Moreover, NCR+ ILCs are enriched in the skin of psoriasis patients (Villanova et al., 2014). In summary, increasing evidence point towards ILCs involvement in several chronic inflammatory diseases.

Figure 6. Development and functional characteristics of ILC subsets.

18

Rheumatoid arthritis (RA) RA is a chronic inflammatory disease that affects joint capsules throughout the body, often in a bilateral and symmetrical manner. The prevalence among the population is approximately 1% and RA is more common among women than men. RA is classified as an autoimmune disease, although it remains unknown which antigen that triggers the immune response to produce autoantibodies i.e. antibodies against endogenous cells. It is stated that 50% of the risk for development of RA is attributable to genetic factors and the main environmental risk is smoking (Scott et al., 2010). Increased serum levels of anti-citrullinated protein antibodies (ACPAs) and rheumatic factors (RFs) are detectable in the majority of patients (Bax et al., 2014).

A joint is largely composed of two cartilage-covered bones surfaces surrounded by a capsule and strengthened by ligaments. The inner surface of the capsule, the synovium, comprises fibroblast-like synoviocytes that produce the synovial fluid filling the joint capsule. In RA, the synovium is infiltrated with inflammatory cells (macrophage-like synoviocytes, fibroblast-like synoviocytes with an altered phenotype, lymphocytes and plasma cells) and becomes hyperplastic, forming a so-called pannus. B and T cells cooperate to promote antibody responses against endogenous proteins and are considered to be major disease mediators. T cells also promote RA independently of B cells mainly through the production of key inflammatory cytokines. Macrophages present in the border between the pannus and the cartilage express pro-inflammatory cytokines (TNF-α, IL-1β), which in turn stimulate the production of MMPs capable of degrading cartilage, which allows the pannus to further invade the cartilage. Bone resorption due to RANKL-mediated osteoclast formation is also seen in active RA (Mölne and Wold, 2007) (Figure 6).

Figure 6. Schematic drawing of a healthy joint (left) and a joint affected by RA (right).

19

Association between periodontitis and RA Despite marked differences in aetiology, the current stage of evidence suggests that there is a clear association between periodontitis and RA. However, the mechanisms underlying the association are yet to be established. According to Kaur et al., three different concepts of interaction have been proposed. The first describes how periodontitis may precede RA through citrullination of proteins by Pophyromonas gingivalis and subsequent production of ACPAs (to be elaborated below). The second suggest that the two diseases share common inflammatory pathways leading to osteoclast activation and vascular damage. The third “two hit model” suggest that periodontitis and RA exacerbate each other through inflammatory mediators (Kaur et al., 2013).

Regardless of these concepts, periodontitis and RA both present with chronic inflammation and tissue destruction. It is suggested that the two diseases are closely related through similarities in the host-mediated inflammatory response (Bartold et al., 2005). More particularly, TNF-α and IL-1β are prominent in the inflammatory process of both conditions and therapeutically blocking their activity has beneficial effects in both periodontitis and RA. Moreover, the inflammatory infiltrate in both conditions is dominated by lymphocytes (Berglundh et al., 2011, Hitchon and El-Gabalawy, 2011). Genetic variances, i.e. polymorphisms, are determinant for the development of both periodontitis and RA. It was early discovered that polymorphisms in human leukocyte antigens (HLA) genes were associated with RA, (Gregersen et al., 1987) which has recently also been demonstrated in periodontitis (Michalowicz et al., 2000). Serum biomarkers are routinely used in the diagnosis of RA and many of the biomarkers of RA are also elevated in periodontitis, e.g. C-reactive protein (CRP), RF and ACPA (Noack et al., 2001, The and Ebersole, 1991, Harvey et al., 2013). In line with this, the periodontal pathogen Porphyromonas gingivalis has been suggested to serve as a common etiologic link between periodontitis and RA due to its ability to enzymatically generate ACPAs via peptidylarginine deiminase (Rosenstein et al., 2004). In addition, the two diseases also share environmental risk factors, e.g. smoking. It has been established that individuals with advanced rheumatoid arthritis are more likely to experience more significant periodontal breakdown compared to their non-diseased counterparts, and vice versa. Interestingly, periodontal treatment is beneficial in terms of reducing the severity of RA (Al-Katma et al., 2007). Moreover, RA patients receiving treatment with TNF-α or IL-6 receptor inhibitors also display a reduction in periodontal inflammation (Kobayashi and Yoshie, 2015). Advocating the possibility of a casual relationship, Ramamurthy et al. could show that induction of experimental arthritis in rats also resulted in increased periodontal breakdown (Ramamurthy et al., 2005).

20

Aims

The overall aim of this thesis was to gain extended knowledge about the role of chemokines in inflammatory disturbed bone remodelling and in periodontitis. Furthermore, the aim was to investigate if the newly discovered ILCs are present in periodontitis and to study the association between periodontitis and RA.

More specifically, the purposes of this thesis were as follows:

• To analyse a spectrum of inflammatory markers in serum from individuals with periodontitis and from periodontally healthy controls and to evaluate their relative importance for disease by using multivariate partial least squares (PLS) modelling. Moreover, we wanted to investigate possible sources of the periodontitis-associated chemokines CCL11 (referred to as eotaxin-1 in paper I) and MCP-1 by analysing their expression in tissue from periodontitis lesions and in human gingival fibroblasts (Paper I).

• To analyse the expression of CCL11 and its cognate receptor CCR3 in vitro in bone cell cultures and in vivo in an inflammatory bone lesion model of mouse parietal bones. In addition, the aim was to investigate the effect of CCL11 on osteoclast formation, migration of osteoclast precursors and bone resorption in vitro (Paper II).

• To explore if RA is preceded by marginal jawbone loss and elevated plasma levels of RANKL (Paper III).

• To further characterise the cellular composition in periodontitis and gingivitis by analysing the presence and relative proportions of innate lymphoid cells (ILCs) in tissue from gingivitis and periodontitis lesions (Paper IV).

21

Methods

This section contains a brief description of methodological approaches. A more detailed description is given within each paper.

Humans All experiments involving humans or human materials in this thesis were approved by the Regional Ethical Review Board in Umeå and conducted according to the principles described in the Declaration of Helsinki and in accordance with Swedish Law on personal data act. Furthermore, a written consent was obtained from all participants.

Gingival explants Gingival papillar explants were obtained from periodontally and systemically healthy individuals under the use of local anaesthesia.

Paper I 84 individuals (>35 years of age), of which 43 were periodontitis cases and 41 were periodontally healthy controls, were recruited at the Specialist Clinic for Periodontology (cases) and the Public Dental Health Clinic (controls) at Norrlands University Hospital. Inclusion and exclusion criteria for each group are defined in the materials and methods section of Paper I. Clinical data regarding periodontal status, general health, medication, and lifestyle variables was recorded and a venous blood sample was collected from each participant into a heparinised tube and stored at the Medical Biobank of Northern Sweden. Three additional periodontitis patients who met the inclusion criteria and underwent periodontal surgery were included in the study for tissue analyses.

Paper III 232 individuals that were symptomatic for RA (cases) and 194 matched controls were identified in the register of the Medical Biobank of Northern Sweden and invited to participate in the study. Study participants were identified by co-analysing the register of patients attending the Department of Rheumatology, University Hospital, Umeå and fulfilling the 1987 ARA classification criteria for RA with those of the Medical Biobank cohorts (Arnett et al., 1988). Both cases and controls were invited, by mail, to participate in the study and asked to fill in a questionnaire on self-assed dental status, smoking habits and contact information for their dental clinic. The selection that ultimately resulted in our final study group of 176 individuals is summarised in the table below. For further details, see paper III, patients and methods.

22

Cases Controls Analysis

Invited to participate 232 194 -

ê ê

Responding 149 143 -

ê ê

X-rays provided by dental clinic 93 83 RF, anti-CCP2, HLA-shared epitope

ê ê

X-rays prior to RA symptom debut

45 45 Bone loss, RANKL

ê ê

Repeated X-rays after RA symptom debut

31 31 Longitudinal evaluation of bone loss

Paper IV: Sample collection Tissue from periodontitis lesions was obtained from individuals undergoing periodontal surgery at the Specialist Clinic for Periodontology at Norrlands University Hospital. Tissue from gingivitis lesions was obtained from individuals undergoing maxillofacial surgery including extraction of one tooth or more at the Specialist Clinic for Oral and Maxillofacial Surgery at Norrlands University Hospital. Tonsils were obtained from individuals undergoing surgical tonsillectomy or tonsillotomy at the Ear, Nose and Throat Clinic at Norrlands University Hospital. Peripheral blood was collected in heparinised tubes.

Mice CsA mice from our own inbred colony were used in all experiments involving mice. Animal care and experiments were approved and conducted in accordance with regulations of the Local Animal-Ethical Committee at Umeå University.

Cell isolation and culture All incubation was conducted at 37°C in humidified air containing 5% CO2.

Human gingival fibroblasts Gingival explants were placed at the bottom of culture dishes with α modification of minimal essential medium (α-MEM) supplemented with 10%

23

foetal bovine serum (FBS), L-glutamine and antibiotics, which will hereinafter be referred to as basic medium. The explants were left untouched for 7-10 days until outgrowth of fibroblasts from the explants could be observed. Medium was changed every other day. When submitted to experiments, the fibroblasts were detached and seeded at a density of 5×104 cells/cm2 in 24-well plates (12-well plates for western blot). After attachment overnight, basic medium was changed and the cells were incubated in the absence (control group) or presence of test substances for different time periods. Cells used in the experiments demonstrated a fibroblastic morphology and were used at passages 5–10. Cell culture supernatants were saved for protein analysis and cell lysates were either subjected to western blot or RNA-isolation and cDNA-synthesis.

Mouse parietal osteoblasts Parietal bones from 2-4 days old mice were dissected aseptically in phosphate-buffered saline (PBS) containing 10% FBS and cells were isolated using the time sequentional collagenase digestion technique as previously described (Boonekamp et al., 1984). In short, cells pooled from digestion fractions 6-10 were placed in a culture bottle containing basic medium and incubated for four days. On day four, the cells were detached and transferred onto 24-well culture plates at a density of 5×104 cells/cm2. After attachment overnight, basic medium was changed and the cells were incubated in the absence (control group) or presence of test substances for different time periods. Cell culture supernatants were saved for protein analysis and cell lysates were subjected to RNA-isolation and cDNA-synthesis. The mouse parietal osteoblasts displayed an osteoblastic phenotype as assessed by their cyclic AMP-responsiveness to PTH, expression of ALP, osteocalcin and BSP in addition to their ability to form mineralised bone noduli (Lerner et al., 1994, Lundberg et al., 1999).

Bone marrow macrophages (BMMs) The femurs and tibiae from 5-8 weeks old male mice were dissected free of adhering soft tissues in PBS. The cartilage/bone ends were cut off and bone marrow cells were flushed out of the marrow space using α-MEM in a syringe with a sterile needle. The marrow cells were collected in basic medium and the erythrocytes were lysed in ice-cold sterile water. All remaining marrow cells were placed on Corning culture dishes in basic medium (referred to as complete medium in paper II) supplemented with M-CSF, in order to stimulate proliferation of cells of the monocyte/macrophage lineage. After 48 h of incubation, non-adherent cells were discarded and remaining adherent macrophages were collected. From this step forward, further procedures depended on the purpose of the BMMs (see osteoclast formation, migration and bone resorption).

24

Pam2-induced inflammation and bone resorption 5 weeks old male CsA mice were injected with Pam2CSK4 (Pam2) or saline (control group) subcutaneously above the parietal bones as described previously (Kassem et al., 2015) and sacrificed after 6 days. The parietal bones were carefully dissected and either committed to homogenisation and RNA isolation or embedding, sectioning and immunohistochemical staining.

Gene expression analysis

RNA isolation and cDNA-synthesis Total RNA from fibroblasts, mouse parietal osteoblasts and BMMs were extracted using the commercially available RNAqueous Micro Kit according to instructions provided by the manufacturer. Whole mouse parietal bones were in a first step homogenised before RNA could be extracted. All extracted total RNA was treated with DNase I to eliminate genomic DNA and quantified using a spectrophotometer. Within the framework of each experiment, equal amounts of RNA (0.3-1.0 µg) were reverse transcribed into single-stranded cDNA with a commercially available kit. Reactions without reverse transcriptase were included as a negative control to verify the absence of genomic DNA in the samples.

RT-qPCR First-strand cDNA mixture was amplified by real-time reverse transcription (RT)- quantitative (q) polymerase chain reaction (PCR). The RT-qPCR analysis was performed using TaqMan Universal PCR Master Mix in combination with custom designed fluorescence labelled probes and oligonucleotide primers or TaqMan Inventoried Gene expression assays. More detailed information regarding primers and probes are listed in the individual papers. The amplifications were detected and analysed using the ABI PRISM 7900 HT Sequence Detection System and associated software. To control and adjust for variances in amplification due to differences in initial mRNA concentrations, endogenous controls were used as internal standards. Human ribosomal protein 13a (hRPL13a) were used for human samples and β-actin or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used for mouse samples. Based on target Ct values and hRPL13a, β-actin or GAPDH values, the relative expression of target mRNAs was calculated using the standard curve method.

25

Protein expression analysis

Serum analysis Collection and handling of blood samples (paper I and III), including fractionation into plasma, serum and buffy coat and storage at -80°C, followed the standardised routines at the Medical Biobank of Northern Sweden, Västerbotten County Council, Sweden. Measurements of inflammatory markers in serum (paper I), including cytokines and chemokines, were performed using Luminex technology. In short, polystyrene beads coated with analyte-specific capture antibodies were added to the samples and bound to the analytes of interest. Next, analyte-specific detection antibodies were added followed by phycoerythrin-conjugated streptavidin. The beads were read on a laser instrument that both classified and determined the amount of bound analyte. The amount of CRP protein was determined using high sensitive immunoturbidimetric assay. In paper III, plasma concentration of the total (free and bound) RANKL was measured using commercially available ELISA kits in accordance with the manufacturer’s protocol. A cut-off value for RANKL positivity was defined using receiver operating characteristic (ROC) curve, yielding a cut off value for positivity. Positivity of anti-CCP2 antibodies, RF and HLA-shared epitope were analysed as previously presented (Boman et al., 2017).

ELISA Total protein concentrations were determined in tissue lysates and cell lysates for western blot assay using bicinchoninic acid (BCA) method in order to adjust for variances in starting protein concentrations. Commercially available enzyme-linked immunosorbent assay (ELISA)-kits were used to assess the amount of intracellular or released CCL11 (referred to as eotaxin-1 in paper I) and MCP-1 protein in cell culture medium or soft tissue lysates. Briefly, samples were added to wells pre-coated with analyte-specific capture antibodies prior to addition of an enzyme-conjugated antibody that also bound to the analyte of interest. Subsequent addition of an enzyme substrate resulted in a change of colour and its intensity, which correlated to the protein concentration, was analysed using a spectrophotometer.

Western Blot Lysates derived from cell cultures were resolved by 10% TRIS-HCl polyacrylamide gel electrophoresis and proteins were transferred onto a nitrocellulose membrane. Protein detection was performed using primary polyclonal rabbit anti-human Iκα antibodies or monoclonal anti-β-actin followed by secondary horseradish peroxidase (HRP)-conjugated anti-rabbit antibodies respectively.

26

Histology and staining Dissected mouse parietal bones were fixed in 4% phosphate-buffered paraformaldehyde, decalcified in 10% ethylenediaminetetraacetic acid (EDTA) and embedded in paraffin. Tissue sections (5 µm-thick) were deparaffinised in xylene and hydrated through a series of graded ethanol-water dilutions. To visualise morphology, sections were stained using the Safranin O/Fast green staining method. Immunohistochemical staining of tissue sections or cells grown on glass cover slips included primary antibodies for TRAP, CCL11 and CCR3. After washing steps, cells/sections were incubated respectively with F-actin or fluorescent secondary antibodies in addition to 4', 6-diamidino-2-phenylindole (DAPI) or HRP-labelled secondary antibodies visualised with 3,3′-diaminobenzidine (DAB) and followed by a nuclear counterstaining with haematoxylin.

For live cell imaging of the internalisation of CCL11 in osteoclast cultures, recombinant mouse CCL11 was labelled with a fluorophore using gravity columns. Before imaging, osteoclasts grown on glass coverslips were transferred to the cell chamber and basic medium was replaced with Dulbecco’s modified eagle medium (DMEM) supplemented with pyruvate before addition of labelled CCL11. Live cell imaging was performed under controlled conditions of 5% CO2 and at 37 °C. A detailed description is found within the methods section of paper II.

TRAP staining of cells on plastic, bone slices and migration plates were performed using the leukocyte acid phosphatase kit according to the protocol provided by the manufacturer. TRAP+ cells with three or more nuclei were considered to be osteoclasts. The number of multinucleated osteoclasts was counted in a light microscope.

Osteoclast formation, migration and bone resorption Depending on the purpose, BMMs derived from tibiae and femurs of mice as described previously, were seeded in culture wells (osteoclast formation and gene expression), culture dishes (migration), cover slips (immunohistochemical staining) or on bone slices (resorption assay) in the presence of basic medium. M-CSF and RANKL were added in order to stimulate osteoclast formation.

To analyse osteoclast differentiation and numbers, the cells were harvested and either stained for TRAP and counted or committed to RNA isolation and cDNA synthesis for gene expression analysis of osteoclast-associated genes, e.g. TRAP and Cathepsin K.

27

BMMs subjected to the migration assay were detached from the culture dishes and added in a globule to the migration filters that were placed on top of wells containing basic medium supplemented with M-CSF and CCL11 or MCP-1. After 5 h of incubation the excess of cells was removed from the top of the filter using a cell-scraper. The filters were detached from the plate, fixed and stained for TRAP. The number of cells stained positive for TRAP was quantified, which was the measure on migrated cells.

In order to analyse the effect of CCL11 on bone resorption, BMMs were seeded on bone slices and recombinant CCL11 was added to the medium. After five and six days all bone slices were TRAP-stained, the number of TRAP positive cells with three or more nuclei (osteoclasts) was counted and the size and area of the resorption pits were measured in reflected light.

Flow cytometry The expression of cell surface markers of the different ILC subsets and RANKL was analysed in single cell suspensions derived from peripheral blood, tonsil or tissue from gingivitis and periodontitis lesions using flow cytometry. Prior to antibody staining, peripheral blood was fractioned and peripheral blood monocytes (PBMC) were collected. Tonsils were directly pressed through a nylon cell-strainer to obtain a single cell suspension. Gingivits and periodontitis tissue was first finely chopped and pre-treated in Collagenase type IV before pressed through the cell strainer. Subsequently, all single cell suspensions were stained with a mix of fluorescent antibodies against cell surface markers of ILC subsets and RANKL. All stainings were performed in the presence of PBS supplemented with human heat-inactivated serum and sodium azide. During staining, the samples were kept on ice and protected from light for 20 minutes.

Radiographic assessment of bone loss Marginal jawbone loss was evaluated in bitewing X-rays reproducing the premolar/molar sections of the jaws. Other types of X-rays, such as apical and panoramic images, were occasionally used to confirm the assessments. For the pre-symptomatic RA individuals (cases) X-rays taken 1 to 7 years before diagnosis of the RA symptoms (referred to as year 0) were selected for evaluation of jawbone loss. For controls, X-rays from the matching age were used. Jawbone loss was scored as present or not present. Presence was scored if a reduction of the marginal bone level corresponding to more than 2/3 of the height of the crown, i.e. 3.7 mm, could be observed. The following were noted for each tooth: (i) intact, i.e., the tooth could be assessed in its entirety and showed no sign of bone loss exceeding 2/3 of the length of the tooth crown; (ii) bone loss (BL), i.e. the tooth displayed bone loss exceeding 2/3 of the length of

28

the tooth crown at any site. If the marginal bone level was not captured in the bite wing radiographs due to severe bone loss it was also scored as BL; (iii) missing. i.e., the tooth was missing or a retained root tip was present; (iv) not assessable, i.e., the tooth could not be evaluated. Total bone loss was defined as the sum of teeth with scored bone loss and lost teeth in proportion of the number of teeth assessable for scoring.

Statistical analysis A detailed description of data handling and statistical analysis performed on data is given within each paper. In paper I, multivariate partial least squares (PLS) modelling was used to detect correlations between variables of an independent block (subject characteristics, periodontitis parameters and inflammatory markers) and having periodontitis or not as a dependent variable. All variables were auto scaled to unit variance and, except for eotaxin and MCP-1, the inflammatory marker concentrations were logarithmically transformed before entered into the model. Furthermore, principal component analysis (PCA) was applied to evaluate clustering by smoking status and inflammatory markers including never and present smokers only. Moreover, Pearson correlation coefficients were calculated between serum concentrations of the inflammatory markers, BMI, and markers for periodontal status with logarithmic transformation of non-normally distributed variables. In paper III, hazard ratios (HR) (95% CI) for being a pre-symptomatic individual for RA by increasing total jawbone loss was calculated from Cox proportional hazard models using time dependent covariate (age) and adjustment for potential confounders (dental care giver and year X-ray was taken). In general, for dichotomized data (Paper I and III), differences between proportions were tested with Chi2-test. For continuous, normally distributed data, differences between means were compared using t-test or one-way analysis of variance (ANOVA) with Levene´s homogeneity test and post hoc Tukey’s test for multiple comparisons. In order to keep type I errors to a minimum, ANOVA was the method of choice when comparing more than two means. When necessary, means were standardised for potential confounders using the general linear model procedure. Non-parametric tests were used to compare rankings when data was non-normally distributed (Paper IV). All experiments were performed at least twice with comparable results. The significance levels were set to P < 0.05 (*), P < 0.01 (**) and P < 0.001 (***).

29

Results and discussion

Measurements of inflammatory markers in serum from individuals with periodontitis and periodontally healthy controls (paper I)

The first task of this project, which also came to serve as a stepping-stone for the majority of the content of this thesis, was to perform a serological screening for a spectrum of inflammatory markers in periodontitis subjects and periodontally healthy controls.

Serum concentrations of all analysed inflammatory markers entered a multivariate PLS modelling analysis together with subject characteristics and periodontitis parameters, which generated a model that separated periodontitis subjects from healthy controls. When analysing each variable separately, we found that elevated concentrations of CRP, CCL11 (referred to as eotaxin-1 in paper I) and MCP-1, in addition to smoking and age, were significantly associated with periodontitis. It has been established that smoking is entailed with an estimated 3-5 fold increased risk of developing periodontitis, which is in accordance with our results (Eke et al., 2015, Papapanou, 1996). On the contrary, the presence of more own teeth and a higher education level were factors associated with periodontal health, as previously described by others (Eke et al., 2015). In accordance with the results from the multivariate model, serum levels of CRP, CCL11 and MCP-1 were significantly higher in periodontitis subjects. Interestingly, the differences in CRP and MCP-1 could only be observed within normal weight subjects, which can be explained by the fact that we detected a positive correlation between BMI and serum levels of CRP and MCP-1.

Elevated serum levels of MCP-1 in periodontitis have previously been described (Gupta et al., 2013, Pradeep et al., 2009), but elevated serum levels of CCL11 have only been detected in a small Brazilian cohort (de Queiroz et al., 2008). Interestingly, decreased levels of CCL11 and MCP-1 in GCF following periodontal therapy have also been demonstrated (Thunell et al., 2010). There are numerous reports on elevated serum levels of CRP in periodontitis (Noack et al., 2001). However, to the best of our knowledge, we are the first to show the explanatory power of CCL11 and MCP-1 in periodontitis using multivariate PLS modelling including known disease confounders, and our results validate and strengthen the association between CCL11 and MCP-1 with periodontitis found by other groups. Our results also support the theory of a systemic increase in inflammatory markers in periodontitis, suggesting that serological screenings

30

could be performed in order to detect periodontitis. Serological biomarkers are used in the diagnosis, prognosis evaluation, and surveillance of other chronic inflammatory diseases. However, there are aggravating factors that need to be taken under consideration when discussing serological biomarkers. In fact, serologic biomarkers are rarely disease specific and they can often only be used as a complement to conventional diagnostic methods. For example, CRP is an acute phase protein that is elevated during several bacterial infections, RA and cancer. Neither are CCL11 nor MCP-1 periodontitis-specific, since they are both also systemically increased in rheumatoid arthritis (Kokkonen et al., 2010, Pavkova Goldbergova et al., 2012), for example. Although our cohort did not present with any other known diseases except for periodontitis, there might be underlying subclinical inflammation present in some of the individuals that might contribute to increased serum inflammatory markers. High BMI represents a condition that may give rise to subclinical inflammation and systemically increased levels of inflammatory mediators (Utsal et al., 2012, Kim et al., 2006). BMI is also associated with an increased risk of periodontitis (Chaffee and Weston, 2010), possibly due to elevated levels of pro-inflammatory cytokines (Boesing et al., 2009). There are several reports on the association between BMI and MCP-1, and the fact that we were able to show increased levels of CRP and MCP-1 in normal weight subjects suggests that both BMI and periodontitis can contribute to a systemic increase in these markers.

Expression and regulation of CCL11 and MCP-1 in periodontitis tissue and in human gingival fibroblasts (Paper I)

Prompted by our initial findings, we wanted to investigate if the periodontitis-associated chemokines CCL11 and MCP-1 could be originating from the periodontitis lesion. Using ELISA, we could identify increased levels of MCP-1 in periodontitis tissue compared to non-inflamed tissue, which is in line with studies conducted by others (Yu et al., 1993). Apart from a study by Lorena et al. that demonstrated the presence of CCL11 in oral mucosa (Lorena et al., 2003), there is no available information regarding CCL11 in periodontitis tissue. Initially, we were unable to detect CCL11 protein in our samples. In follow-up analyses performed after publication of these findings, we were able to detect increased levels of CCL11 in periodontitis, using an ELISA kit with a wider detection range (unpublished data). The gingival fibroblast is the predominant cell type in gingival connective tissue. In addition to constructing collagen, gingival fibroblasts are able to act during inflammatory conditions by regulating the host’s immune response due to production and release of pro-inflammatory cytokines, chemokines and prostaglandins. Hence, we analysed the expression of CCL11 and MCP-1 in human gingival fibroblasts and could demonstrate

31

increased expression of MCP-1 and CCL11 mRNA and protein in the cells when they were challenged with TNF-α and IL-1β, key mediators of periodontal inflammation. Other research groups have also demonstrated expression of CCL11 and MCP-1 in human gingival fibroblasts (Hosokawa et al., 2013, Sawada et al., 2013), which strengthens our results, but we could add that the observed effects were both time- and concentration dependent. Addition of three different pharmacological inhibitors of the intracellular pathway of NF-κB, which is a pathway commonly used by TNF α and IL-1β (Baldwin, 1996), resulted in a considerable decrease in CCL11 and MCP-1 mRNA expression. Thus, it was concluded that NF-κB signalling mediates expression of CCL11 and MCP-1 by TNF-α and IL-1β in human gingival fibroblasts.

In summary, the gathered results in paper I demonstrate that periodontitis is characterised by systemically (serum) and locally (gingiva) increased levels of CCL11 and MCP-1, and that the chemokines are produced by gingival fibroblasts in response to TNF-α and IL-1β through activation of the intracellular signalling pathway NF-κB.

Expression of CCL11 and CCR3 during inflammatory conditions in vivo in an inflammatory bone lesion model of mouse parietal bones and in vitro in bone resident cells (Paper II)

Chemokines are suggested to be involved in several aspects of bone metabolism including the recruitment of leukocytes of the monocyte/macrophage lineage from the bone marrow and later the formation of bone resorbing osteoclasts. Thus, we wanted to further investigate the role of CCL11 and MCP-1 in inflammatory disturbed bone remodelling. It has previously been demonstrated that MCP-1 is expressed in osteoblasts (Williams et al., 1992) and during osseous inflammation (Rahimi et al., 1995). Furthermore, Li et al. showed that MCP-1 mediates the migration and subsequent fusion of osteoclast precursors (Li et al., 2007). In a mouse model of experimental periodontitis induced by Porphyromonas gingivalis, mice that were treated with a CCR2-antagonist were less prone to develop marginal jawbone loss (Barros et al., 2011). On the contrary, little is known regarding the function of CCL11 and its receptor CCR3 in bone metabolism, which prompted us to analyse CCL11 and CCR3 expression in an in vivo inflammatory bone lesion model and in vitro in bone resident cells.

First we analysed the CCL11 expression in vivo. Subcutaneous injection of toll-like receptor agonist Pam2 above the parietal bones of 5 weeks old mice, according to Kassem et al. (Kassem et al., 2015) induced a potent inflammatory response and extensive bone resorption in the parietal bones, which in

32

morphologically stained sections of bone was seen as an infiltration of inflammatory cells, presence of osteoclasts and disruption of normal bone structure. Mice of the control group, which received saline injection, maintained their normal bone structure. The prevailing inflammatory process was confirmed by mRNA analysis on homogenates of parietal bone derived from parallel experiments, by which we could demonstrate increased expression of TNF-α and IL-1β mRNA in bones of Pam2-treated mice. Next we investigated if Pam2-induced inflammation and bone resorption resulted in local production of CCL11 using immunohistochemical methods. In the saline-treated group as well as in the Pam2-treated group, CCL11 protein was detected in all periosteal osteoblasts and in a few cells in the marrow space. Multinucleated cells adjacent to bone surfaces in samples from Pam2-treated mice also stained positive for CCL11, and by TRAP staining we could verify that these cells were indeed osteoclasts. By quantitatively assessing mRNA levels in homogenates of parietal bones, we could confirm increased levels of CCL11 in Pam2-injected mice compared to saline control.

In our in vivo model, we identified CCL11 positivity in mouse parietal osteoblasts and also in osteoclasts, which prompted us to further investigate their respective mRNA and released protein expression of CCL11. Mouse parietal osteoblasts demonstrated increased CCL11 expression, both mRNA and released protein, in response to TNF-α and IL-β. Apart from a study by Röhner et al., that reported enhanced CCL11 protein expression in human osteoblasts due to exposure by polyhexanide (Rohner et al., 2015), constitutive and inflammatory stimulated expression of CCL11 in osteoblasts have not previously been investigated. Furthermore, we analysed CCL11 mRNA expression in BMM cultures, to which M-CSF and RANKL were added in order to promote osteoclast differentiation. Most surprisingly, CCL11 mRNA expression was undetectable in RANKL-derived osteoclasts, despite the presence of osteoclasts with marked positivity for CCL11 in the in vivo model. In accordance with our findings, there are no previous reports on CCL11 expression by osteoclasts. We hypothesised that the CCL11 that was evident on osteoclasts in the immunohistochemically stained sections, might be receptor-bound and not expressed by the osteoclasts themselves.

Due to the fact that CCL11 is capable of interacting with several CC chemokine receptors (CCR2, CCR3 and CCR5), we proceeded to investigate if osteoclast differentiation affected the mRNA expression levels of these receptors. CCR3 is the main receptor for CCL11 (Daugherty et al., 1996) and interestingly, CCR3 mRNA expression was significantly upregulated during osteoclast differentiation at 2 and 3 days of culture. In contrast, the mRNA levels of CCR2 and CCR5 were significantly downregulated. These results are in accordance with previous reports demonstrating that CCR3 is expressed during RANKL-

33

mediated osteoclastogenesis and that both CCR2 and CCR5 are suppressed (Lean et al., 2002, Kominsky et al., 2008, Matsubara et al., 2012). We could also confirm that the expression of CCR2, CCR3 and CCR5 in human osteoclasts displayed a similar pattern (unpublished data). Taken together, these observations strengthened our theory of receptor-bound CCL11 in osteoclasts, which in turn encouraged further studies on receptor-ligand interaction.

CCL11 binding and effects on osteoclast formation, migration of osteoclast precursors and bone resorption (Paper II)

In order to visualise the subcellular localisation of CCR3, RANKL-derived osteoclasts were co-immunolabelled for CCR3 and F-actin. The latter is a multi-functional protein that forms the microfilaments constituting the osteoclasts cytoskeleton (Woo et al., 2002). By including F-actin, we could depict the osteoclast morphology in its entirety and use it as a guide when localising CCR3. Confocal microscopy revealed that CCR3 was localised to protrusions and ruffled borders in mononuclear preosteoclasts. In mature multinuclear osteoclasts, CCR3 was detected at the tip of thin actin-rich protrusions polarised to specific areas of the cells, in addition to showing a homogeneous intracellular distribution. The localisation of CCR3 on osteoclasts has to our knowledge not been described previously. The polarised localisation of CCR3 suggested that it might be important to sense the environment, facilitate motility and guide migration of osteoclasts to resorption sites (Kanehisa et al., 1990). In mast cells, CCL11 causes migration through interaction with CCR3, downstream activation of Rac and ERK and subsequent cytoskeletal rearrangements (Woo et al., 2002). In order for CCL11 to interact with CCR3 on osteoclasts, it would require direct contact and interaction with CCR3. To study the possible interaction between CCL11 and CCR3 in osteoclasts we first wanted to investigate if we could detect co-localisation of CCL11 and CCR3 on cell surfaces. We immunolabelled CCL11 and CCR3 in parietal bone sections from mice treated with Pam2, in which osteoclasts were present on bone surfaces. We found that osteoclasts and osteoblasts stained positive for both CCL11 and CCR3. In osteoclasts, it was apparent that the localisation of CCL11 overlapped with that of CCR3 in discrete areas at the subcellular level. This finding suggests that there is a possible interaction between CCL11 and CCR3, which in turn might be important for osteoclast function.

The RANKL-mediated upregulation of CCR3 during osteoclast differentiation, together with the evident co-localisation of CCL11 and CCR3 in osteoclasts from sites displaying inflammatory disturbed bone remodelling, raised the question whether CCL11 binding could stimulate osteoclast formation, migration of osteoclast precursors and bone resorption. Using live imaging, we visualised the

34

interplay between fluorescently labelled CCL11 that was added to osteoclasts cultured on glass, and it was revealed that CCL11 was efficiently internalised within osteoclasts. Addition of CCL11 to RANKL in BMM cultures had no effect on osteoclastogenesis as assessed by the number of osteoclasts. To investigate chemotactic migration of osteoclast precursors, we used a migration assay in which osteoclast precursors were allowed to migrate towards CCL11. These experiments demonstrated, for the first time, that CCL11 significantly stimulated the migration of osteoclast precursors. In year 2000, Votta et al. investigated the effects of eotaxin-2 on the migration of osteoclast precursors and found that it had no chemotactic effect (Votta et al., 2000). Eotaxin-2/CCL24 is despite similarities in nomenclature, not closely related to CCL11/eotaxin-1, thus our results are not comparable. In fact, CCL11 and MCP-1 are more closely related, showing 66% amino acid sequence homology. MCP-1, a chemokine with known effects on the migration of preosteoclasts, (Li et al., 2007) was included in the assay as a positive control. Ultimately we wanted to study if CCL11, in addition to the chemotactic effects on osteoclast precursors, could also enhance bone resorption or modulate the resorption pattern. The addition of CCL11 to RANKL in BMMs cultured on bone slices resulted in an altered pattern of bone resorption, more particularly an increase in total resorption area.

In summary, the highly interesting findings of paper II describe newly discovered properties of CCL11 in inflammatory disturbed bone remodelling. We were able to demonstrate co-localisation of CCL11 and CCR3 on osteoclasts, and describe important effector functions of CCL11 on osteoclast precursors and osteoclasts. However, our findings can only indicate, but not prove, that CCL11 exerts its effects through binding to CCR3. It would seem logical that the CCL11/CCR3 axis is indeed activated since CCL11 shows very high affinity to CCR3 compared to other chemokine receptors. In osteoarthritis and rheumatoid arthritis, both chronic inflammatory diseases with disturbed bone remodelling, it has been demonstrated that CCR3 is upregulated and participate the inflammatory process by activating fibroblast-like synoviocytes (Chang et al., 2016, Liu et al., 2017). Further studies involving CCR3-deficient mice are necessary in order to elucidate the causative effects of the interaction between CCL11 and CCR3.

Radiographically assessed jawbone loss in pre-symptomatic RA cases and matched controls (Paper III)

Despite marked differences in aetiology, periodontitis and RA share several common features and it is suggested that they are inter-related. The association between the two diseases is evident, but the extent and type of association have

35

not yet been determined (Araujo et al., 2015), presumably due to the lack of prospective studies performed within the field.

In paper III, we aspired to assess whether periodontitis, displayed as marginal jawbone loss, preceded the onset of symptoms of RA and if plasma RANKL-levels were related to jawbone loss in individuals who subsequently develop RA.

The basic study group consisted of 176 individuals from which dental radiographs could be obtained. There were no differences in the proportions of men/women, never/ever smokers and dental care provider as well as mean age and birth year between the groups. Among the pre-symptomatic cases in the basic study group, 79% were RF positive and 75% were anti-CCP2 positive, which is in accordance with other studies (Verheul et al., 2015, Abdel-Nasser et al., 2008). Out of these 176 individuals, we could identify 45 future RA cases (referred to as pre-symptomatic individuals) with X-rays available prior to RA symptom debut and 45 controls matched for sex, age and smoking status. Marginal jawbone loss was evaluated in X-rays from the 45 matched pairs.

Although our study material was limited, univariate analyses revealed that the mean number of intact teeth (number of teeth where jawbone loss did not exceed 2/3 of the height of the tooth crown) was significantly lower in pre-symptomatic individuals compared to controls. In accordance, pre-symptomatic individuals also had a higher proportion of teeth with jawbone loss. Stratification for smoking habits revealed that the differences were only seen in subjects who reported always being non-smokers. Further analyses between developing RA symptoms and degree of jawbone loss were performed in all 45 matched pairs. After adjustment for potential confounders, we detected a positive association between developing RA and jawbone loss among the non-smokers. The fact that no association between marginal jawbone loss and RA was observed among smokers, which corresponds with a study by Potikuri (Potikuri et al., 2012), is highly interesting. Smoking is a common environmental factor associated with both diseases, which could explain why the association was only evident in non-smokers. Furthermore, the longitudinal analyses of bone loss revealed that the increment in mean bone loss was higher in pre-symptomatic individuals than in matched controls.

Due to the fact that we investigated marginal jawbone loss before onset of RA, our results are not comparable to cross-sectional studies performed within the field that have been able to demonstrate that RA patients display higher prevalence of periodontitis (Kasser et al., 1997, Mercado et al., 2001, Pischon et al., 2008, Garib and Qaradaxi, 2011, de Smit et al., 2012, Scher et al., 2012, Potikuri et al., 2012, Joseph et al., 2013). In line with our findings, Hashimoto et al. demonstrated that among patients with arthralgia, those who also had

36

periodontitis presented with higher arthritis activity and were more likely to receive treatment with methotrexate (Hashimoto et al., 2015). There are studies with contradictory results (Demmer et al., 2011, Eriksson et al., 2016), which could be due to several aggravating factors, including diversities in disease classifications, cohort sizes and the fact that treatment of RA could affect periodontal health (Kobayashi and Yoshie, 2015). A prospective study by Arkema and collaborators found no association between severe periodontitis, estimated by the history of periodontal surgery and/or tooth loss, and the risk of RA (Arkema et al., 2010). The different results between the study by Arkema et al. and our study could be due to the fact that only individuals with severe periodontitis undergo periodontal surgery and because of that, the results are not applicable to the general population. Estimation of marginal jawbone loss gives a more nuanced description of the severity of the periodontal disease. However, evaluation of jawbone loss is not sufficient in order to diagnose and classify periodontal disease, which is an indisputable weak point of our study. Information regarding periodontal status (gingival pocket depth and gingival bleeding) was either lacking our inconsistent in the available material, thus we were not able to diagnose or classify periodontitis. However, it is reasonable to believe that the marginal jawbone loss was caused by past or present periodontitis.

Herein, we found that RANKL positive pre-symptomatic individuals had a significantly higher extent of marginal jawbone loss and we could also show a modest correlation between plasma concentrations of RANKL and jawbone loss. The fact that RANKL positive pre-symptomatic individuals had more severe marginal jawbone loss indicates that there may be a sub-group of pre-symptomatic individuals who share a pathological pathway of inflammatory disturbed bone remodelling with individuals with severe periodontitis. It must be clarified that our investigation is not sufficient to determine if marginal jawbone loss is the source of RANKL or if RANKL is causative for the development of RA in pre-symptomatic individuals. Furthermore, the concentration of RANKL was significantly higher in pre-symptomatic individuals positive for ACPA or RF but we were unable to detect a correlation between RF positivity and jawbone loss in any of the groups. The highest values of jawbone loss were detected in pre-symptomatic individuals that were both ACPA and RANKL positive. ACPAs have previously been described in periodontitis (de Pablo et al., 2014) and the periodontal pathogen Porphyromonas gingivalis is capable of producing citrullinated proteins and generate ACPAs. Thus, it is hypothesised that ACPAs may act as a possible link between periodontitis and RA (Rosenstein et al., 2004, Krishnamurthy et al., 2016). Moreover, ACPAs are shown to potentiate RANKL-induced osteoclast formation and bone resorption (Krishnamurthy et al., 2016) it is possible that

37

our findings on high jawbone loss RANKL-positive pre-symptomatic individuals could be driven by ACPAs

The observations made in paper III, although interpretation should be careful due to our reported study limitations, support the view of an association between periodontitis and RA but additional research is needed to determine whether there is a causative link between periodontitis and RA or if the diseases are simply the outcome of a pattern of inflammatory disturbed bone remodelling present in susceptible individuals.

Presence and relative proportions of ILCs and their expression of RANKL in gingivitis and periodontitis (Paper IV)

The increasing evidence of ILCs as important actors in the pathogenesis of chronic inflammatory diseases at mucosal surfaces together with the fact that there is a deficient picture of the cellular composition in gingivitis and periodontitis lesions prompted us to investigate the presence of ILCs in these conditions.

We applied the classification of ILC subsets as defined by Hazenberg and Spits (Hazenberg and Spits, 2014), and by using this approach, cells that did not express cell surface markers for T-cells (CD3-) and lineage (CD1a- CD3-, CD14-, CD19- CD34- CD94- CD123- FcεR1α- TCRαβ- TCRγδ- BDCA2-) but being CD127+ and CD161+ were selected for further characterisation. From this step forward, the cells were divided into the different ILC subsets: ILC1 (CRTH2- CD117-), ILC2 (CRTH2+), NCR- ILC3 (CRTH2- CD117+ NKp44-) and NCR+ ILC3 (CRTH2-

CD117+ NKp44+).

We detected twice as many ILCs in periodontitis as in gingivitis, as could be expected since it has been demonstrated that the inflammatory infiltrate in gingivitis is not as extensive as in periodontitis (Berglundh et al., 2011). Only a few leukocytes and sporadic ILCs were detected in healthy gingiva. The numeric difference between the groups did not reach statistical significance, which can possibly be explained by the relatively few tissue samples included in the analysis in combination with large dispersions in ILC numbers. Surgery directed towards gingivitis sites in adults was rarely performed, which made it problematic to gather enough gingivitis tissue. In addition, all tissue samples had to be handled and analysed immediately following surgery, which resulted in extensive laboratory work. ILCs constitute a recently identified group of leukocytes and to the best of our knowledge, these cells have not been described in gingivitis or periodontitis previously. In 2013, Gladiator and colleagues demonstrated that IL-17 producing ILCs are present in oral mucosa and mediate

38

host defence mechanisms against fungal infection (Gladiator et al., 2013). In addition, Dutzan et al. recently demonstrated the presence of ILCs in healthy gingiva (Dutzan et al., 2016). Our findings on the presence of ILCs in gingivitis and periodontitis provide a new insight into the cellular composition of these conditions.

Furthermore, we analysed the proportions of different ILC subsets in gingivitis and in periodontitis. The presence of all four subsets could be detected to varying degrees in both conditions, whereas ILC1s constituted the majority of the ILC compartment followed by NCR- ILC3s. The proportions of ILC2s and NCR+ ILC3s were almost negligible. Interestingly, in periodontitis there was a proportional and numerical increase in ILC1s (a 2.5 fold increase in the number of cells compared to gingivitis). ILC1s are enriched in chronic inflammation of the gut (Bernink et al., 2013), and contribute to disease pathogenesis through dysregulated production of IFN-γ in response to IL-12. Accordingly, an antibody targeting IL-12 might be beneficial for Chron’s patients that do not respond to anti-TNF treatment, possibly through reduced levels of IFN- γ (Mannon et al., 2004). The same antibody has also been suggested to be beneficial in the treatment of periodontitis (Moutsopoulos et al., 2017). Periodontitis is reflected by increased levels of IFN-γ in serum, saliva and GCF (Andrukhov et al., 2011, Isaza-Guzman et al., 2015, Papathanasiou et al., 2014). However, the role of IFN-γ in periodontitis and inflammatory disturbed bone remodelling is not fully elicudated and the studies performed within the field are contradictory. It is suggested that IFN-γ might stimulate osteoclast formation and bone resorption, either directly (Madyastha et al., 2000) or by inducing TNF-α or RANKL (Gao et al., 2007), which in relation to our findings propose a function of ILC1 in periodontitis. In contrast, IFN-γ has also been demonstrated to inhibit osteoclastic formation (Kamolmatyakul et al., 2001), and it is speculated if the effect of IFN-γ on osteoclastogenesis might be biphasic (Cheng et al., 2012).

Since ILCs, particularly ILC1s were more abundant in periodontitis than in gingivitis we hypothesised that they might possess unknown osteotropic properties. Hence, we analysed the expression of RANKL on different ILC subsets and detected a RANKL expression that was almost exclusively limited to ILC1s. This finding adds a new property to ILCs and suggests that they be involved in inflammatory disturbed bone remodelling throughout the body. However, studies involving a larger number of samples and preferably mechanistic studies are required to elucidate the role of ILCs in gingivitis and periodontitis.

39

Conclusions

The gathered observations presented in this thesis indicate that the chemokines CCL11 and MCP-1 may be involved in inflammatory disturbed bone remodelling and in the pathogenesis of periodontitis. The finding of ILCs in gingivitis and periodontitis provides a new insight into the cellular composition of these conditions and into the field of oral immunology. Furthermore, the results also support the theory of an association between periodontitis and RA. In summary, the findings supporting our conclusions are:

• Serum levels of the chemokines CCL11 and MCP-1 are identified as clearly associated with periodontitis. Periodontitis is reflected by systemically (serum) and locally (gingiva) increased levels of CCL11 and MCP-1. Primary human gingival fibroblasts display strongly increased NF-κB-dependent expression of CCL11 and MCP-1 mRNA and protein in response to TNF-α and IL-1β

• CCL11 is expressed by osteoblasts during inflammatory conditions in vitro and in vivo. Osteoclasts do not express CCL11, but its cognate receptor CCR3 is significantly upregulated during osteoclast differentiation. Furthermore, CCL11 co-localises with CCR3 on the surface of osteoclasts and exogenous CCL11 added to cell cultures is internalised within osteoclasts. CCL11 has a positive effect on the migration of osteoclast precursors and stimulates osteoclastic bone resorption in vitro.

• ILCs are more abundant in periodontitis than in gingivitis and the majority of ILCs, in both conditions, belong to the ILC1 subset. Furthermore, RANKL expression is observed almost exclusively on ILC1s compared to other subsets.

• Marginal jawbone loss precedes the onset of symptoms of RA, particularly in non-smoking pre-symptomatic individuals. Moreover, RANKL positive pre-symptomatic RA-individuals display a significantly higher degree of marginal jawbone loss, particularly in addition to ACPA positivity.

40

Acknowledgements

My years as a PhD-student have truly been the most exciting and evolving time of my life and I would like to express my sincere gratitude to all who have helped me to accomplish this work. In particular, I want to thank the following people:

First and foremost, my biggest thanks go to my supervisor, associate professor Pernilla Lundberg. I want to thank you for giving me the opportunity to be a part of your research group, for providing me with new challenges that have lead me to develop and for the fact that you always find a way of being present, despite your packed schedule. Owing to your excellent guidance, I have not once doubted the quality or the feasibility of this project. As a PhD-student, I could not have hoped for a better supervisor!

Cecilia Koskinen Holm, my co-supervisor and roommate, thank you for always making time for my questions regardless of how busy you are, for being honest and constructive, and for teaching me the importance of daily exercise in the staircase. Your way of handling osteoclasts and your effectiveness are impressive!

Py Palmqvist, my co-supervisor. Thank you for your guidance through these years and for always being considerate, helpful and enthusiastic. Your sense of humour and your storytelling-skills are one of a kind!

The rest of the research group: Rima Sulniute, histology guru, thank you for your practical guidance in the laboratory, for all the times you saved the day and for contributing to this thesis with invaluable results and beautiful stainings. Your positive spirit and your way of making everything seem effortless are impressive! Sara Engman, my room-mate and dear friend, thank you for your supporting attitude, our almost daily coffee expeditions to Kahls, the laugther we have shared and all the miscalculations we have made together during these years. I hope I get to be your colleague for many, many years to come! Catrine Isehed, thank your for showing sincere interest in my work and for your collection of biopsies. Best of luck with your thesis! Sussie Lindquist and Michaela Eklund, our newest members, I hope we will do great research together! Johanna Lång, thank you for helping me set up the migration assay and for great company.

41

Anders Johansson, my research education examiner. I admire the fact that you never turn down a challenge, whether it’s a venous blood sample or an obstacle course. Thank you for always being helpful and understanding and for all nice discussions in the coffee room.

Kristina Lejon, thank you for educating me in flow cytometry and more importantly, for teaching me the fine art of coloring graphs. Your positive attitude and true passion for science has made our collaboration a true pleasure.

Professor Ingegerd Johansson, thank you for for teaching me the basics of more advanced statistical methods in a most pedagogic way, I really appreciate it.

Professor Solbritt Rantapää Dahlqvist, thank you for sharing your great knowledge about rheumatology with me.

Professor Ulf Lerner, thank you for your thoughtful insights.

All co-workers and co-authors are acknowledged, but in particular I want to thank Ingrid Boström, Inger Lundgren and Chrissie Roth for teaching me the basic methods, for untiringly answering my lab-related questions and for contributing to a good working environment. Anita Lie, thank you for teaching me how to culture osteoclasts and for your invaluable contributions to this thesis. I really admire your accuracy and determination! Elisabeth Granström, Jan Oscarsson and Mark Lindholm, thank you for bringing new life to the department of Molecular Periodontology and for nice companionship. Britt-Inger Nordell, thank you for your help with organising patient journals and X-rays.

Mats Sjöström, Carola Höglund-Åberg, Peder Wilhelmsson, Hanna Löf and Uday Najim, thank you for your collection of biopsies.

To all administrators, thank you for all your assistance and special thanks to Sari Korva for helping me arrange for my part-time studies.

To all senior lecturers at THU, thank you for supporting me in my part-time studies and making it possible for me to graduate with my classmates.

To the Stillesjö family, thank you for being such good friends to my family and me, and a special thanks to Sara for always giving me good advice in PhD-related matters.

42

Sara Rönnberger and Jonas Carlsson, thank you for being my devoted audience and for your invaluable friendship, I miss your presence in Umeå.

My mother Elisabeth and my father Mats. Thank you for letting me practice “learning by doing” and for always supporting me in all my projects, no matter big or small, realistic or unrealistic. You provided me with a great self-confidence, for which I am ever grateful. My brother Jonas, thank you for all the Sunday dinners we had together while you lived in Umeå and for sharing my interest in science.

To my grandparents Runhild and Per-Olov, Erik and my late grandmother Inga-Britt. Thank you for spoiling me with delicious “palt” and for making me feel like the most special person in the whole world.

My parents-in-law, Bengt and Mirella, thank you for welcoming me into your family and for the way you care about Ralf.

Kalle, my husband and best friend, thank you for being my biggest supporter, for always finding solutions to my IT-related problems, for the tremendous care you show towards my family and friends and for being the kind-hearted person that you are. My beloved son Ralf, thank you for giving me the most amazing experience of my life!

43

References

Abdel-Nasser, A. M., Mahmoud, M. H., El Mansoury, T. M. & Osman, A. M. 2008. Anti-CCP2 is an adjunct to, not a surrogate for, rheumatoid factor in the diagnosis of rheumatoid arthritis: diagnostic utility of anti-CCP2 antibodies in Egyptian patients with rheumatoid arthritis. Scand J Rheumatol, 37, 329-36. 10.1080/03009740802116208

Agrawal, M., Arora, S., Li, J., Rahmani, R., Sun, L., Steinlauf, A. F., Mechanick, J. I. & Zaidi, M. 2011. Bone, inflammation, and inflammatory bowel disease. Curr Osteoporos Rep, 9, 251-7. 10.1007/s11914-011-0077-9

Al-Katma, M. K., Bissada, N. F., Bordeaux, J. M., Sue, J. & Askari, A. D. 2007. Control of periodontal infection reduces the severity of active rheumatoid arthritis. J Clin Rheumatol, 13, 134-7. 10.1097/RHU.0b013e3180690616

Anderson, H. C. 2003. Matrix vesicles and calcification. Curr Rheumatol Rep, 5, 222-6.

Andrukhov, O., Ulm, C., Reischl, H., Nguyen, P. Q., Matejka, M. & Rausch-Fan, X. 2011. Serum cytokine levels in periodontitis patients in relation to the bacterial load. J Periodontol, 82, 885-92. 10.1902/jop.2010.100425

Araujo, V. M., Melo, I. M. & Lima, V. 2015. Relationship between Periodontitis and Rheumatoid Arthritis: Review of the Literature. Mediators Inflamm, 2015, 259074. 10.1155/2015/259074

Arkema, E. V., Karlson, E. W. & Costenbader, K. H. 2010. A prospective study of periodontal disease and risk of rheumatoid arthritis. J Rheumatol, 37, 1800-4. 10.3899/jrheum.091398

Arnett, F. C., Edworthy, S. M., Bloch, D. A., Mcshane, D. J., Fries, J. F., Cooper, N. S., Healey, L. A., Kaplan, S. R., Liang, M. H., Luthra, H. S. & Et Al. 1988. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum, 31, 315-24.

Aubin, J. E. & Heersche, J. N. M. 2000. Osteoprogenitor cell differentiation to mature bone-forming osteoblasts. Drug Development Research, 49, 206-215. Doi 10.1002/(Sici)1098-2299(200003)49:3<206::Aid-Ddr11>3.0.Co;2-G

Baldwin, A. S., Jr. 1996. The NF-kappa B and I kappa B proteins: new discoveries and insights. Annu Rev Immunol, 14, 649-83. 10.1146/annurev.immunol.14.1.649

Bar-Ephraim, Y. E. & Mebius, R. E. 2016. Innate lymphoid cells in secondary lymphoid organs. Immunol Rev, 271, 185-99. 10.1111/imr.12407

44

Barros, S. P., Arce, R. M., Galloway, P., Lawter, R. & Offenbacher, S. 2011. Therapeutic effect of a topical CCR2 antagonist on induced alveolar bone loss in mice. J Periodontal Res, 46, 246-51. 10.1111/j.1600-0765.2010.01340.x

Bartold, P. M., Marshall, R. I. & Haynes, D. R. 2005. Periodontitis and rheumatoid arthritis: a review. J Periodontol, 76, 2066-74. 10.1902/jop.2005.76.11-S.2066

Bartold, P. M., Walsh, L. J. & Narayanan, A. S. 2000. Molecular and cell biology of the gingiva. Periodontol 2000, 24, 28-55.

Bax, M., Huizinga, T. W. & Toes, R. E. 2014. The pathogenic potential of autoreactive antibodies in rheumatoid arthritis. Semin Immunopathol, 36, 313-25. 10.1007/s00281-014-0429-5

Bendre, M. S., Margulies, A. G., Walser, B., Akel, N. S., Bhattacharrya, S., Skinner, R. A., Swain, F., Ramani, V., Mohammad, K. S., Wessner, L. L., Martinez, A., Guise, T. A., Chirgwin, J. M., Gaddy, D. & Suva, L. J. 2005. Tumor-derived interleukin-8 stimulates osteolysis independent of the receptor activator of nuclear factor-kappaB ligand pathway. Cancer Res, 65, 11001-9. 10.1158/0008-5472.CAN-05-2630

Bendre, M. S., Montague, D. C., Peery, T., Akel, N. S., Gaddy, D. & 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.

Berglundh, T., Zitzmann, N. U. & Donati, M. 2011. Are peri-implantitis lesions different from periodontitis lesions? J Clin Periodontol, 38 Suppl 11, 188-202. 10.1111/j.1600-051X.2010.01672.x

Bernink, J. H., Peters, C. P., Munneke, M., Te Velde, A. A., Meijer, S. L., Weijer, K., Hreggvidsdottir, H. S., Heinsbroek, S. E., Legrand, N., Buskens, C. J., Bemelman, W. A., Mjosberg, J. M. & Spits, H. 2013. Human type 1 innate lymphoid cells accumulate in inflamed mucosal tissues. Nat Immunol, 14, 221-9. 10.1038/ni.2534

Binder, N. B., Niederreiter, B., Hoffmann, O., Stange, R., Pap, T., Stulnig, T. M., Mack, M., Erben, R. G., Smolen, J. S. & Redlich, K. 2009. Estrogen-dependent and C-C chemokine receptor-2-dependent pathways determine osteoclast behavior in osteoporosis. Nat Med, 15, 417-24. 10.1038/nm.1945

Boesing, F., Patino, J. S., Da Silva, V. R. & Moreira, E. A. 2009. The interface between obesity and periodontitis with emphasis on oxidative stress and inflammatory response. Obes Rev, 10, 290-7. 10.1111/j.1467-789X.2008.00555.x

45

Boman, A., Kokkonen, H., Arlestig, L., Berglin, E. & Rantapaa-Dahlqvist, S. 2017. Receptor activator of nuclear factor kappa-B ligand (RANKL) but not sclerostin or gene polymorphisms is related to joint destruction in early rheumatoid arthritis. Clin Rheumatol, 36, 1005-1012. 10.1007/s10067-017-3570-4

Bonewald, L. F. & Mundy, G. R. 1990. Role of transforming growth factor-beta in bone remodeling. Clin Orthop Relat Res, 261-76.

Boonekamp, P. M., Hekkelman, J. W., Hamilton, J. W., Cohn, D. V. & Jilka, R. L. 1984. Effect of Culture on the Hormone Responsiveness of Bone-Cells Isolated by an Improved Sequential Digestion Procedure. Proceedings of the Koninklijke Nederlandse Akademie Van Wetenschappen Series B-Palaeontology Geology Physics Chemistry Anthropology, 87, 371-381.

Boyce, B. F. 2013. Advances in osteoclast biology reveal potential new drug targets and new roles for osteoclasts. J Bone Miner Res, 28, 711-22. 10.1002/jbmr.1885

Boyce, B. F., Li, P., Yao, Z., Zhang, Q., Badell, I. R., Schwarz, E. M., O'keefe, R. J. & Xing, L. 2005. TNF-alpha and pathologic bone resorption. Keio J Med, 54, 127-31.

Boyle, W. J., Simonet, W. S. & Lacey, D. L. 2003. Osteoclast differentiation and activation. Nature, 423, 337-42. 10.1038/nature01658

Burr, D. B. 2002. Targeted and nontargeted remodeling. Bone, 30, 2-4.

Chaffee, B. W. & Weston, S. J. 2010. Association between chronic periodontal disease and obesity: a systematic review and meta-analysis. J Periodontol, 81, 1708-24. 10.1902/jop.2010.100321

Chang, X., Shen, J., Yang, H., Xu, Y., Gao, W., Wang, J., Zhang, H. & He, S. 2016. Upregulated expression of CCR3 in osteoarthritis and CCR3 mediated activation of fibroblast-like synoviocytes. Cytokine, 77, 211-9. 10.1016/j.cyto.2015.09.012

Cheng, J., Liu, J., Shi, Z., Jules, J., Xu, D., Luo, S., Wei, S. & Feng, X. 2012. Molecular mechanisms of the biphasic effects of interferon-gamma on osteoclastogenesis. J Interferon Cytokine Res, 32, 34-45. 10.1089/jir.2011.0019

Clarke, B. 2008. Normal bone anatomy and physiology. Clin J Am Soc Nephrol, 3 Suppl 3, S131-9. 10.2215/CJN.04151206

Daugherty, B. L., Siciliano, S. J., Demartino, J. A., Malkowitz, L., Sirotina, A. & Springer, M. S. 1996. Cloning, expression, and characterization of the human eosinophil eotaxin receptor. J Exp Med, 183, 2349-54.

46

De Pablo, P., Dietrich, T., Chapple, I. L., Milward, M., Chowdhury, M., Charles, P. J., Buckley, C. D. & Venables, P. J. 2014. The autoantibody repertoire in periodontitis: a role in the induction of autoimmunity to citrullinated proteins in rheumatoid arthritis? Ann Rheum Dis, 73, 580-6. 10.1136/annrheumdis-2012-202701

De Queiroz, A. C., Taba, M., Jr., O'connell, P. A., Da Nobrega, P. B., Costa, P. P., Kawata, V. K., Trevisan, G. L., Novaes, A. B., Jr., De Souza, S. L., Palioto, D. B. & Grisi, M. F. 2008. Inflammation markers in healthy and periodontitis patients: a preliminary data screening. Braz Dent J, 19, 3-8.

De Smit, M., Westra, J., Vissink, A., Doornbos-Van Der Meer, B., Brouwer, E. & Van Winkelhoff, A. J. 2012. Periodontitis in established rheumatoid arthritis patients: a cross-sectional clinical, microbiological and serological study. Arthritis Res Ther, 14, R222. 10.1186/ar4061

Delaisse, J. M. 2014. The reversal phase of the bone-remodeling cycle: cellular prerequisites for coupling resorption and formation. Bonekey Rep, 3, 561. 10.1038/bonekey.2014.56

Demmer, R. T., Molitor, J. A., Jacobs, D. R., Jr. & Michalowicz, B. S. 2011. Periodontal disease, tooth loss and incident rheumatoid arthritis: results from the First National Health and Nutrition Examination Survey and its epidemiological follow-up study. J Clin Periodontol, 38, 998-1006. 10.1111/j.1600-051X.2011.01776.x

Dewhirst, F. E., Stashenko, P. P., Mole, J. E. & Tsurumachi, T. 1985. Purification and partial sequence of human osteoclast-activating factor: identity with interleukin 1 beta. J Immunol, 135, 2562-8.

Dietrich, J. W., Goodson, J. M. & Raisz, L. G. 1975. Stimulation of bone resorption by various prostaglandins in organ culture. Prostaglandins, 10, 231-40.

Dougall, W. C., Glaccum, M., Charrier, K., Rohrbach, K., Brasel, K., De Smedt, T., Daro, E., Smith, J., Tometsko, M. E., Maliszewski, C. R., Armstrong, A., Shen, V., Bain, S., Cosman, D., Anderson, D., Morrissey, P. J., Peschon, J. J. & Schuh, J. 1999. RANK is essential for osteoclast and lymph node development. Genes Dev, 13, 2412-24.

Dutzan, N., Konkel, J. E., Greenwell-Wild, T. & Moutsopoulos, N. M. 2016. Characterization of the human immune cell network at the gingival barrier. Mucosal Immunol, 9, 1163-72. 10.1038/mi.2015.136

Eberl, G., Colonna, M., Di Santo, J. P. & Mckenzie, A. N. 2015. Innate lymphoid cells. Innate lymphoid cells: a new paradigm in immunology. Science, 348, aaa6566. 10.1126/science.aaa6566

47

Eke, P. I., Dye, B. A., Wei, L., Slade, G. D., Thornton-Evans, G. O., Borgnakke, W. S., Taylor, G. W., Page, R. C., Beck, J. D. & Genco, R. J. 2015. Update on Prevalence of Periodontitis in Adults in the United States: NHANES 2009 to 2012. J Periodontol, 86, 611-22. 10.1902/jop.2015.140520

Eriksson, K., Nise, L., Kats, A., Luttropp, E., Catrina, A. I., Askling, J., Jansson, L., Alfredsson, L., Klareskog, L., Lundberg, K. & Yucel-Lindberg, T. 2016. Prevalence of Periodontitis in Patients with Established Rheumatoid Arthritis: A Swedish Population Based Case-Control Study. PLoS One, 11, e0155956. 10.1371/journal.pone.0155956

Felix, R., Hofstetter, W., Wetterwald, A., Cecchini, M. G. & Fleisch, H. 1994. Role of colony-stimulating factor-1 in bone metabolism. J Cell Biochem, 55, 340-9. 10.1002/jcb.240550311

Galliera, E., Locati, M., Mantovani, A. & Corsi, M. M. 2008. Chemokines and bone remodeling. Int J Immunopathol Pharmacol, 21, 485-91. 10.1177/039463200802100301

Gao, Y., Grassi, F., Ryan, M. R., Terauchi, M., Page, K., Yang, X., Weitzmann, M. N. & Pacifici, R. 2007. IFN-gamma stimulates osteoclast formation and bone loss in vivo via antigen-driven T cell activation. J Clin Invest, 117, 122-32. 10.1172/JCI30074

Garib, B. T. & Qaradaxi, S. S. 2011. Temporomandibular joint problems and periodontal condition in rheumatoid arthritis patients in relation to their rheumatologic status. J Oral Maxillofac Surg, 69, 2971-8. 10.1016/j.joms.2011.02.131

Gladiator, A., Wangler, N., Trautwein-Weidner, K. & Leibundgut-Landmann, S. 2013. Cutting edge: IL-17-secreting innate lymphoid cells are essential for host defense against fungal infection. J Immunol, 190, 521-5. 10.4049/jimmunol.1202924

Gowen, M., Wood, D. D., Ihrie, E. J., Mcguire, M. K. & Russell, R. G. 1983. An interleukin 1 like factor stimulates bone resorption in vitro. Nature, 306, 378-80.

Gregersen, P. K., Silver, J. & Winchester, R. J. 1987. The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum, 30, 1205-13.

Gupta, M., Chaturvedi, R. & Jain, A. 2013. Role of monocyte chemoattractant protein-1 (MCP-1) as an immune-diagnostic biomarker in the pathogenesis of chronic periodontal disease. Cytokine, 61, 892-7. 10.1016/j.cyto.2012.12.012

Harvey, G. P., Fitzsimmons, T. R., Dhamarpatni, A. A., Marchant, C., Haynes, D. R. & Bartold, P. M. 2013. Expression of peptidylarginine deiminase-2 and -4,

48

citrullinated proteins and anti-citrullinated protein antibodies in human gingiva. J Periodontal Res, 48, 252-61. 10.1111/jre.12002

Hashimoto, M., Yamazaki, T., Hamaguchi, M., Morimoto, T., Yamori, M., Asai, K., Isobe, Y., Furu, M., Ito, H., Fujii, T., Terao, C., Mori, M., Matsuo, T., Yoshitomi, H., Yamamoto, K., Yamamoto, W., Bessho, K. & Mimori, T. 2015. Periodontitis and Porphyromonas gingivalis in preclinical stage of arthritis patients. PLoS One, 10, e0122121. 10.1371/journal.pone.0122121

Hassell, T. M. & Harris, E. L. 1995. Genetic influences in caries and periodontal diseases. Crit Rev Oral Biol Med, 6, 319-42.

Hazenberg, M. D. & Spits, H. 2014. Human innate lymphoid cells. Blood, 124, 700-9. 10.1182/blood-2013-11-427781

Hitchon, C. A. & El-Gabalawy, H. S. 2011. The synovium in rheumatoid arthritis. Open Rheumatol J, 5, 107-14. 10.2174/1874312901105010107

Hock, J. M., Centrella, M. & Canalis, E. 1988. Insulin-like growth factor I has independent effects on bone matrix formation and cell replication. Endocrinology, 122, 254-60. 10.1210/endo-122-1-254

Hofbauer, L. C., Lacey, D. L., Dunstan, C. R., Spelsberg, T. C., Riggs, B. L. & Khosla, S. 1999. Interleukin-1beta and tumor necrosis factor-alpha, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells. Bone, 25, 255-9.

Horwood, N. J., Elliott, J., Martin, T. J. & Gillespie, M. T. 1998. Osteotropic agents regulate the expression of osteoclast differentiation factor and osteoprotegerin in osteoblastic stromal cells. Endocrinology, 139, 4743-6. 10.1210/endo.139.11.6433

Hoshino, A., Iimura, T., Ueha, S., Hanada, S., Maruoka, Y., Mayahara, M., Suzuki, K., Imai, T., Ito, M., Manome, Y., Yasuhara, M., Kirino, T., Yamaguchi, A., Matsushima, K. & Yamamoto, K. 2010. Deficiency of chemokine receptor CCR1 causes osteopenia due to impaired functions of osteoclasts and osteoblasts. J Biol Chem, 285, 28826-37. 10.1074/jbc.M109.099424

Hosokawa, Y., Hosokawa, I., Shindo, S., Ozaki, K. & Matsuo, T. 2013. (-)-Epigallocatechin-3-gallate inhibits CC chemokine ligand 11 production in human gingival fibroblasts. Cell Physiol Biochem, 31, 960-7. 10.1159/000350114

Huang, J. C., Sakata, T., Pfleger, L. L., Bencsik, M., Halloran, B. P., Bikle, D. D. & Nissenson, R. A. 2004. PTH differentially regulates expression of RANKL and OPG. J Bone Miner Res, 19, 235-44. 10.1359/JBMR.0301226

49

Isaza-Guzman, D. M., Cardona-Velez, N., Gaviria-Correa, D. E., Martinez-Pabon, M. C., Castano-Granada, M. C. & Tobon-Arroyave, S. I. 2015. Association study between salivary levels of interferon (IFN)-gamma, interleukin (IL)-17, IL-21, and IL-22 with chronic periodontitis. Arch Oral Biol, 60, 91-9. 10.1016/j.archoralbio.2014.09.002

Ishii, M., Egen, J. G., Klauschen, F., Meier-Schellersheim, M., Saeki, Y., Vacher, J., Proia, R. L. & Germain, R. N. 2009. Sphingosine-1-phosphate mobilizes osteoclast precursors and regulates bone homeostasis. Nature, 458, 524-8. 10.1038/nature07713

Ishimi, Y., Miyaura, C., Jin, C. H., Akatsu, T., Abe, E., Nakamura, Y., Yamaguchi, A., Yoshiki, S., Matsuda, T., Hirano, T. & Et Al. 1990. IL-6 is produced by osteoblasts and induces bone resorption. J Immunol, 145, 3297-303.

Jilka, R. L. 2003. Biology of the basic multicellular unit and the pathophysiology of osteoporosis. Med Pediatr Oncol, 41, 182-5. 10.1002/mpo.10334

Jonsson, D., Nebel, D., Bratthall, G. & Nilsson, B. O. 2011. The human periodontal ligament cell: a fibroblast-like cell acting as an immune cell. J Periodontal Res, 46, 153-7. 10.1111/j.1600-0765.2010.01331.x

Joseph, R., Rajappan, S., Nath, S. G. & Paul, B. J. 2013. Association between chronic periodontitis and rheumatoid arthritis: a hospital-based case-control study. Rheumatol Int, 33, 103-9. 10.1007/s00296-011-2284-1

Kamolmatyakul, S., Chen, W. & Li, Y. P. 2001. Interferon-gamma down-regulates gene expression of cathepsin K in osteoclasts and inhibits osteoclast formation. J Dent Res, 80, 351-5. 10.1177/00220345010800011001

Kanehisa, J. & Heersche, J. N. 1988. Osteoclastic bone resorption: in vitro analysis of the rate of resorption and migration of individual osteoclasts. Bone, 9, 73-9.

Kanehisa, J., Yamanaka, T., Doi, S., Turksen, K., Heersche, J. N., Aubin, J. E. & Takeuchi, H. 1990. A band of F-actin containing podosomes is involved in bone resorption by osteoclasts. Bone, 11, 287-93.

Kassebaum, N. J., Bernabe, E., Dahiya, M., Bhandari, B., Murray, C. J. & Marcenes, W. 2014. Global burden of severe periodontitis in 1990-2010: a systematic review and meta-regression. J Dent Res, 93, 1045-53. 10.1177/0022034514552491

Kassem, A., Henning, P., Lundberg, P., Souza, P. P., Lindholm, C. & Lerner, U. H. 2015. Porphyromonas gingivalis Stimulates Bone Resorption by Enhancing RANKL (Receptor Activator of NF-kappaB Ligand) through Activation of Toll-

50

like Receptor 2 in Osteoblasts. J Biol Chem, 290, 20147-58. 10.1074/jbc.M115.655787

Kasser, U. R., Gleissner, C., Dehne, F., Michel, A., Willershausen-Zonnchen, B. & Bolten, W. W. 1997. Risk for periodontal disease in patients with longstanding rheumatoid arthritis. Arthritis Rheum, 40, 2248-51. 10.1002/1529-0131(199712)40:12&lt;2248::AID-ART20&gt;3.0.CO;2-W

Kaur, S., White, S. & Bartold, P. M. 2013. Periodontal disease and rheumatoid arthritis: a systematic review. J Dent Res, 92, 399-408. 10.1177/0022034513483142

Kikuta, J., Iwai, K., Saeki, Y. & Ishii, M. 2011. S1P-targeted therapy for elderly rheumatoid arthritis patients with osteoporosis. Rheumatol Int, 31, 967-9. 10.1007/s00296-010-1634-8

Kim, C. S., Park, H. S., Kawada, T., Kim, J. H., Lim, D., Hubbard, N. E., Kwon, B. S., Erickson, K. L. & Yu, R. 2006. Circulating levels of MCP-1 and IL-8 are elevated in human obese subjects and associated with obesity-related parameters. Int J Obes (Lond), 30, 1347-55. 10.1038/sj.ijo.0803259

Kim, M. S., Day, C. J. & Morrison, N. A. 2005. MCP-1 is induced by receptor activator of nuclear factor-{kappa}B ligand, promotes human osteoclast fusion, and rescues granulocyte macrophage colony-stimulating factor suppression of osteoclast formation. J Biol Chem, 280, 16163-9. 10.1074/jbc.M412713200

Kinane, D. F., Berglundh, T. & Lindhe, J. 2008. Pathogenesis of Periodontitis. In: Lindhe, J., Lang, N. P. & Karring, T. (eds.) Clinical Periodontology and Implant Dentistry. 5 ed. Oxford: Blackwell Munksgaard.

Kobayashi, K., Takahashi, N., Jimi, E., Udagawa, N., Takami, M., Kotake, S., Nakagawa, N., Kinosaki, M., Yamaguchi, K., Shima, N., Yasuda, H., Morinaga, T., Higashio, K., Martin, T. J. & Suda, T. 2000. Tumor necrosis factor alpha stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKL-RANK interaction. J Exp Med, 191, 275-86.

Kobayashi, T. & Yoshie, H. 2015. Host Responses in the Link Between Periodontitis and Rheumatoid Arthritis. Curr Oral Health Rep, 2, 1-8. 10.1007/s40496-014-0039-2

Kokkonen, H., Soderstrom, I., Rocklov, J., Hallmans, G., Lejon, K. & Rantapaa Dahlqvist, S. 2010. Up-regulation of cytokines and chemokines predates the onset of rheumatoid arthritis. Arthritis Rheum, 62, 383-91. 10.1002/art.27186

Kominsky, S. L., Abdelmagid, S. M., Doucet, M., Brady, K. & Weber, K. L. 2008. Macrophage inflammatory protein-1 delta: a novel osteoclast stimulating factor secreted by renal cell carcinoma bone metastasis. Cancer Res, 68, 1261-6. 10.1158/0008-5472.CAN-07-6122

51

Komori, T., Yagi, H., Nomura, S., Yamaguchi, A., Sasaki, K., Deguchi, K., Shimizu, Y., Bronson, R. T., Gao, Y. H., Inada, M., Sato, M., Okamoto, R., Kitamura, Y., Yoshiki, S. & Kishimoto, T. 1997. Targeted disruption of Cbfa1 results in a complete lack of bone formation owing to maturational arrest of osteoblasts. Cell, 89, 755-64.

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. & Penninger, J. M. 1999. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature, 397, 315-23. 10.1038/16852

Konya, V. & Mjosberg, J. 2015. Innate lymphoid cells in graft-versus-host disease. Am J Transplant, 15, 2795-801. 10.1111/ajt.13394

Konya, V. & Mjosberg, J. 2016. Lipid mediators as regulators of human ILC2 function in allergic diseases. Immunol Lett, 179, 36-42. 10.1016/j.imlet.2016.07.006

Krishnamurthy, A., Joshua, V., Haj Hensvold, A., Jin, T., Sun, M., Vivar, N., Ytterberg, A. J., Engstrom, M., Fernandes-Cerqueira, C., Amara, K., Magnusson, M., Wigerblad, G., Kato, J., Jimenez-Andrade, J. M., Tyson, K., Rapecki, S., Lundberg, K., Catrina, S. B., Jakobsson, P. J., Svensson, C., Malmstrom, V., Klareskog, L., Wahamaa, H. & Catrina, A. I. 2016. Identification of a novel chemokine-dependent molecular mechanism underlying rheumatoid arthritis-associated autoantibody-mediated bone loss. Ann Rheum Dis, 75, 721-9. 10.1136/annrheumdis-2015-208093

Lean, J. M., Murphy, C., Fuller, K. & Chambers, T. J. 2002. CCL9/MIP-1gamma and its receptor CCR1 are the major chemokine ligand/receptor species expressed by osteoclasts. J Cell Biochem, 87, 386-93. 10.1002/jcb.10319

Lerner, U. H., Lundberg, P., Ransjo, M., Persson, P. & Hakanson, R. 1994. Helodermin, helospectin, and PACAP stimulate cyclic AMP formation in intact bone, isolated osteoblasts, and osteoblastic cell lines. Calcif Tissue Int, 54, 284-9.

Lerner, U. H. & Persson, E. 2008. Osteotropic effects by the neuropeptides calcitonin gene-related peptide, substance P and vasoactive intestinal peptide. J Musculoskelet Neuronal Interact, 8, 154-65.

Li, J., Sarosi, I., Yan, X. Q., Morony, S., Capparelli, C., Tan, H. L., Mccabe, S., Elliott, R., Scully, S., Van, G., Kaufman, S., Juan, S. C., Sun, Y., Tarpley, J., Martin, L., Christensen, K., Mccabe, J., Kostenuik, P., Hsu, H., Fletcher, F., Dunstan, C. R., Lacey, D. L. & Boyle, W. J. 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, 97, 1566-71.

52

Li, X., Qin, L., Bergenstock, M., Bevelock, L. M., Novack, D. V. & Partridge, N. C. 2007. Parathyroid hormone stimulates osteoblastic expression of MCP-1 to recruit and increase the fusion of pre/osteoclasts. J Biol Chem, 282, 33098-106. 10.1074/jbc.M611781200

Lian, J. B. & Stein, G. S. 1995. Development of the osteoblast phenotype: molecular mechanisms mediating osteoblast growth and differentiation. Iowa Orthop J, 15, 118-40.

Lindhe, J., Karring, T. & Araujo, M. 2008. The Anatomy of Periodontal Tissues. In: Lindhe, J., Lang, N. P. & Karring, T. (eds.) Clinical Periodontology and Implant Dentistry. 5 ed. Oxford: Blackwell Munksgaard.

Linkhart, T. A., Mohan, S. & Baylink, D. J. 1996. Growth factors for bone growth and repair: IGF, TGF beta and BMP. Bone, 19, 1S-12S.

Liu, X., Zhang, H., Chang, X., Shen, J., Zheng, W., Xu, Y., Wang, J., Gao, W. & He, S. 2017. Upregulated expression of CCR3 in rheumatoid arthritis and CCR3-dependent activation of fibroblast-like synoviocytes. Cell Biol Toxicol, 33, 15-26. 10.1007/s10565-016-9356-7

Liu, Y. C., Kao, Y. T., Huang, W. K., Lin, K. Y., Wu, S. C., Hsu, S. C., Schuyler, S. C., Li, L. Y., Leigh Lu, F. & Lu, J. 2014. CCL5/RANTES is important for inducing osteogenesis of human mesenchymal stem cells and is regulated by dexamethasone. Biosci Trends, 8, 138-43.

Locklin, R. M., Oreffo, R. O. & Triffitt, J. T. 1999. Effects of TGFbeta and bFGF on the differentiation of human bone marrow stromal fibroblasts. Cell Biol Int, 23, 185-94. 10.1006/cbir.1998.0338

Loe, H., Anerud, A., Boysen, H. & Morrison, E. 1986. Natural history of periodontal disease in man. Rapid, moderate and no loss of attachment in Sri Lankan laborers 14 to 46 years of age. J Clin Periodontol, 13, 431-45.

Lorena, S. C., Oliveira, D. T., Dorta, R. G., Landman, G. & Kowalski, L. P. 2003. Eotaxin expression in oral squamous cell carcinomas with and without tumour associated tissue eosinophilia. Oral Dis, 9, 279-83.

Lundberg, P., Bostrom, I., Mukohyama, H., Bjurholm, A., Smans, K. & Lerner, U. H. 1999. Neuro-hormonal control of bone metabolism: vasoactive intestinal peptide stimulates alkaline phosphatase activity and mRNA expression in mouse calvarial osteoblasts as well as calcium accumulation mineralized bone nodules. Regul Pept, 85, 47-58.

Madyastha, P. R., Yang, S., Ries, W. L. & Key, L. L., Jr. 2000. IFN-gamma enhances osteoclast generation in cultures of peripheral blood from osteopetrotic patients and normalizes superoxide production. J Interferon Cytokine Res, 20, 645-52. 10.1089/107999000414826

53

Mannon, P. J., Fuss, I. J., Mayer, L., Elson, C. O., Sandborn, W. J., Present, D., Dolin, B., Goodman, N., Groden, C., Hornung, R. L., Quezado, M., Yang, Z., Neurath, M. F., Salfeld, J., Veldman, G. M., Schwertschlag, U., Strober, W. & Anti, I. L. C. S. D. S. G. 2004. Anti-interleukin-12 antibody for active Crohn's disease. N Engl J Med, 351, 2069-79. 10.1056/NEJMoa033402

Martin, T. J. & Sims, N. A. 2005. Osteoclast-derived activity in the coupling of bone formation to resorption. Trends Mol Med, 11, 76-81. 10.1016/j.molmed.2004.12.004

Matsubara, R., Kukita, T., Ichigi, Y., Takigawa, I., Qu, P. F., Funakubo, N., Miyamoto, H., Nonaka, K. & Kukita, A. 2012. Characterization and identification of subpopulations of mononuclear preosteoclasts induced by TNF-alpha in combination with TGF-beta in rats. PLoS One, 7, e47930. 10.1371/journal.pone.0047930

Mercado, F. B., Marshall, R. I., Klestov, A. C. & Bartold, P. M. 2001. Relationship between rheumatoid arthritis and periodontitis. J Periodontol, 72, 779-87. 10.1902/jop.2001.72.6.779

Michalowicz, B. S., Aeppli, D. P., Kuba, R. K., Bereuter, J. E., Conry, J. P., Segal, N. L., Bouchard, T. J., Jr. & Pihlstrom, B. L. 1991. A twin study of genetic variation in proportional radiographic alveolar bone height. J Dent Res, 70, 1431-5. 10.1177/00220345910700110701

Michalowicz, B. S., Diehl, S. R., Gunsolley, J. C., Sparks, B. S., Brooks, C. N., Koertge, T. E., Califano, J. V., Burmeister, J. A. & Schenkein, H. A. 2000. Evidence of a substantial genetic basis for risk of adult periodontitis. J Periodontol, 71, 1699-707. 10.1902/jop.2000.71.11.1699

Miller, S. C., De Saint-Georges, L., Bowman, B. M. & Jee, W. S. 1989. Bone lining cells: structure and function. Scanning Microsc, 3, 953-60; discussion 960-1.

Mjosberg, J. & Spits, H. 2016. Human Innate Lymphoid Cells. J Allergy Clin Immunol. 10.1016/j.jaci.2016.09.009

Mölne, J. & Wold, A. 2007. Kronisk inflammation. In: Olsson, A. E.-M. (ed.) Inflammation. 1 ed. Stockholm: Författarna och Liber AB.

Monticelli, L. A., Sonnenberg, G. F., Abt, M. C., Alenghat, T., Ziegler, C. G., Doering, T. A., Angelosanto, J. M., Laidlaw, B. J., Yang, C. Y., Sathaliyawala, T., Kubota, M., Turner, D., Diamond, J. M., Goldrath, A. W., Farber, D. L., Collman, R. G., Wherry, E. J. & Artis, D. 2011. Innate lymphoid cells promote lung-tissue homeostasis after infection with influenza virus. Nat Immunol, 12, 1045-54. 10.1031/ni.2131

54

Moutsopoulos, N. M., Zerbe, C. S., Wild, T., Dutzan, N., Brenchley, L., Dipasquale, G., Uzel, G., Axelrod, K. C., Lisco, A., Notarangelo, L. D., Hajishengallis, G., Notarangelo, L. D. & Holland, S. M. 2017. Interleukin-12 and Interleukin-23 Blockade in Leukocyte Adhesion Deficiency Type 1. N Engl J Med, 376, 1141-1146. 10.1056/NEJMoa1612197

Nakashima, K., Zhou, X., Kunkel, G., Zhang, Z., Deng, J. M., Behringer, R. R. & De Crombrugghe, B. 2002. The novel zinc finger-containing transcription factor osterix is required for osteoblast differentiation and bone formation. Cell, 108, 17-29.

Nesbitt, S., Nesbit, A., Helfrich, M. & Horton, M. 1993. Biochemical characterization of human osteoclast integrins. Osteoclasts express alpha v beta 3, alpha 2 beta 1, and alpha v beta 1 integrins. J Biol Chem, 268, 16737-45.

Noack, B., Genco, R. J., Trevisan, M., Grossi, S., Zambon, J. J. & De Nardin, E. 2001. Periodontal infections contribute to elevated systemic C-reactive protein level. J Periodontol, 72, 1221-7. 10.1902/jop.2000.72.9.1221

Oba, Y., Lee, J. W., Ehrlich, L. A., Chung, H. Y., Jelinek, D. F., Callander, N. S., Horuk, R., Choi, S. J. & Roodman, G. D. 2005. MIP-1alpha utilizes both CCR1 and CCR5 to induce osteoclast formation and increase adhesion of myeloma cells to marrow stromal cells. Exp Hematol, 33, 272-8. 10.1016/j.exphem.2004.11.015

Ott, S. M. 2002. Histomorphometric Analysis of Bone Remodelling. In: Bilezikian, J. P., Raisz, L. G. & Rodan, G. A. (eds.) Principles of Bone Biology. 2 ed. San Diego: Academic Press.

Otto, F., Thornell, A. P., Crompton, T., Denzel, A., Gilmour, K. C., Rosewell, I. R., Stamp, G. W., Beddington, R. S., Mundlos, S., Olsen, B. R., Selby, P. B. & Owen, M. J. 1997. Cbfa1, a candidate gene for cleidocranial dysplasia syndrome, is essential for osteoblast differentiation and bone development. Cell, 89, 765-71.

Page, R. C. & Kornman, K. S. 1997. The pathogenesis of human periodontitis: an introduction. Periodontol 2000, 14, 9-11.

Page, R. C., Offenbacher, S., Schroeder, H. E., Seymour, G. J. & Kornman, K. S. 1997. Advances in the pathogenesis of periodontitis: summary of developments, clinical implications and future directions. Periodontol 2000, 14, 216-48.

Palmer, R. & Soory, M. 2008. Modifying Factors. In: Lindhe, J., Lang, N. P. & Karring, T. (eds.) Clinical Periodontology and Implant Dentistry. 5 ed. Oxford: Blackwell Munksgaard.

Papapanou, P. N. 1996. Periodontal diseases: epidemiology. Ann Periodontol, 1, 1-36. 10.1902/annals.1996.1.1.1

55

Papathanasiou, E., Teles, F., Griffin, T., Arguello, E., Finkelman, M., Hanley, J. & Theoharides, T. C. 2014. Gingival crevicular fluid levels of interferon-gamma, but not interleukin-4 or -33 or thymic stromal lymphopoietin, are increased in inflamed sites in patients with periodontal disease. J Periodontal Res, 49, 55-61. 10.1111/jre.12078

Parfitt, A. M. 2002. Targeted and nontargeted bone remodeling: relationship to basic multicellular unit origination and progression. Bone, 30, 5-7.

Pavkova Goldbergova, M., Lipkova, J., Pavek, N., Gatterova, J., Vasku, A., Soucek, M. & Nemec, P. 2012. RANTES, MCP-1 chemokines and factors describing rheumatoid arthritis. Mol Immunol, 52, 273-8. 10.1016/j.molimm.2012.06.006

Pischon, N., Pischon, T., Kroger, J., Gulmez, E., Kleber, B. M., Bernimoulin, J. P., Landau, H., Brinkmann, P. G., Schlattmann, P., Zernicke, J., Buttgereit, F. & Detert, J. 2008. Association among rheumatoid arthritis, oral hygiene, and periodontitis. J Periodontol, 79, 979-86. 10.1902/jop.2008.070501

Ponomaryov, T., Peled, A., Petit, I., Taichman, R. S., Habler, L., Sandbank, J., Arenzana-Seisdedos, F., Magerus, A., Caruz, A., Fujii, N., Nagler, A., Lahav, M., Szyper-Kravitz, M., Zipori, D. & Lapidot, T. 2000. Induction of the chemokine stromal-derived factor-1 following DNA damage improves human stem cell function. J Clin Invest, 106, 1331-9. 10.1172/JCI10329

Potikuri, D., Dannana, K. C., Kanchinadam, S., Agrawal, S., Kancharla, A., Rajasekhar, L., Pothuraju, S. & Gumdal, N. 2012. Periodontal disease is significantly higher in non-smoking treatment-naive rheumatoid arthritis patients: results from a case-control study. Ann Rheum Dis, 71, 1541-4. 10.1136/annrheumdis-2011-200380

Pradeep, A. R., Daisy, H. & Hadge, P. 2009. Serum levels of monocyte chemoattractant protein-1 in periodontal health and disease. Cytokine, 47, 77-81. 10.1016/j.cyto.2009.05.012

Rahimi, P., Wang, C. Y., Stashenko, P., Lee, S. K., Lorenzo, J. A. & Graves, D. T. 1995. Monocyte chemoattractant protein-1 expression and monocyte recruitment in osseous inflammation in the mouse. Endocrinology, 136, 2752-9. 10.1210/endo.136.6.7750500

Ramamurthy, N. S., Greenwald, R. A., Celiker, M. Y. & Shi, E. Y. 2005. Experimental arthritis in rats induces biomarkers of periodontitis which are ameliorated by gene therapy with tissue inhibitor of matrix metalloproteinases. J Periodontol, 76, 229-33. 10.1902/jop.2005.76.2.229

Rodan, G. A. & Martin, T. J. 1982. Role of osteoblasts in hormonal control of bone resorption - a hypothesis. Calcif Tissue Int, 34, 311.

56

Rohner, E., Hoff, P., Gaber, T., Lang, A., Voros, P., Buttgereit, F., Perka, C., Windisch, C. & Matziolis, G. 2015. Cytokine expression in human osteoblasts after antiseptic treatment: a comparative study between polyhexanide and chlorhexidine. J Invest Surg, 28, 1-7. 10.3109/08941939.2014.941445

Rosenstein, E. D., Greenwald, R. A., Kushner, L. J. & Weissmann, G. 2004. Hypothesis: the humoral immune response to oral bacteria provides a stimulus for the development of rheumatoid arthritis. Inflammation, 28, 311-8. 10.1007/s10753-004-6641-z

Saika, M., Inoue, D., Kido, S. & Matsumoto, T. 2001. 17beta-estradiol stimulates expression of osteoprotegerin by a mouse stromal cell line, ST-2, via estrogen receptor-alpha. Endocrinology, 142, 2205-12. 10.1210/endo.142.6.8220

Sawada, S., Chosa, N., Ishisaki, A. & Naruishi, K. 2013. Enhancement of gingival inflammation induced by synergism of IL-1beta and IL-6. Biomed Res, 34, 31-40.

Schaffler, M. B., Cheung, W. Y., Majeska, R. & Kennedy, O. 2014. Osteocytes: master orchestrators of bone. Calcif Tissue Int, 94, 5-24. 10.1007/s00223-013-9790-y

Scher, J. U., Ubeda, C., Equinda, M., Khanin, R., Buischi, Y., Viale, A., Lipuma, L., Attur, M., Pillinger, M. H., Weissmann, G., Littman, D. R., Pamer, E. G., Bretz, W. A. & Abramson, S. B. 2012. Periodontal disease and the oral microbiota in new-onset rheumatoid arthritis. Arthritis Rheum, 64, 3083-94. 10.1002/art.34539

Scott, D. L., Wolfe, F. & Huizinga, T. W. 2010. Rheumatoid arthritis. Lancet, 376, 1094-108. 10.1016/S0140-6736(10)60826-4

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, J., Davy, E., Bucay, N., Renshaw-Gegg, L., Hughes, T. M., Hill, D., Pattison, W., Campbell, P., Sander, S., Van, G., Tarpley, J., Derby, P., Lee, R. & Boyle, W. J. 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell, 89, 309-19.

Sodek, J. & Mckee, M. D. 2000. Molecular and cellular biology of alveolar bone. Periodontol 2000, 24, 99-126.

Sonnenberg, G. F. 2014. Regulation of intestinal health and disease by innate lymphoid cells. Int Immunol, 26, 501-7. 10.1093/intimm/dxu052

Sonnenberg, G. F., Fouser, L. A. & Artis, D. 2011. Border patrol: regulation of immunity, inflammation and tissue homeostasis at barrier surfaces by IL-22. Nat Immunol, 12, 383-90. 10.1038/ni.2025

57

Souza, P. P. & Lerner, U. H. 2013. The role of cytokines in inflammatory bone loss. Immunol Invest, 42, 555-622. 10.3109/08820139.2013.822766

Suda, T., Takahashi, N., Udagawa, N., Jimi, E., Gillespie, M. T. & Martin, T. J. 1999. Modulation of osteoclast differentiation and function by the new members of the tumor necrosis factor receptor and ligand families. Endocr Rev, 20, 345-57. 10.1210/edrv.20.3.0367

Sul, O. J., Ke, K., Kim, W. K., Kim, S. H., Lee, S. C., Kim, H. J., Kim, S. Y., Suh, J. H. & Choi, H. S. 2012. Absence of MCP-1 leads to elevated bone mass via impaired actin ring formation. J Cell Physiol, 227, 1619-27. 10.1002/jcp.22879

Takahashi, N., Udagawa, N. & Suda, T. 2014. Vitamin D endocrine system and osteoclasts. Bonekey Rep, 3, 495. 10.1038/bonekey.2013.229

Takayanagi, H., Kim, S., Koga, T., Nishina, H., Isshiki, M., Yoshida, H., Saiura, A., Isobe, M., Yokochi, T., Inoue, J., Wagner, E. F., Mak, T. W., Kodama, T. & Taniguchi, T. 2002. Induction and activation of the transcription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts. Dev Cell, 3, 889-901.

Tanaka, S., Takahashi, N., Udagawa, N., Tamura, T., Akatsu, T., Stanley, E. R., Kurokawa, T. & Suda, T. 1993. Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin Invest, 91, 257-63. 10.1172/JCI116179

Tatsumi, S., Ishii, K., Amizuka, N., Li, M., Kobayashi, T., Kohno, K., Ito, M., Takeshita, S. & Ikeda, K. 2007. Targeted ablation of osteocytes induces osteoporosis with defective mechanotransduction. Cell Metab, 5, 464-75. 10.1016/j.cmet.2007.05.001

Thaiss, C. A., Zmora, N., Levy, M. & Elinav, E. 2016. The microbiome and innate immunity. Nature, 535, 65-74. 10.1038/nature18847

The, J. & Ebersole, J. L. 1991. Rheumatoid factor (RF) distribution in periodontal disease. J Clin Immunol, 11, 132-42.

Thunell, D. H., Tymkiw, K. D., Johnson, G. K., Joly, S., Burnell, K. K., Cavanaugh, J. E., Brogden, K. A. & Guthmiller, J. M. 2010. A multiplex immunoassay demonstrates reductions in gingival crevicular fluid cytokines following initial periodontal therapy. J Periodontal Res, 45, 148-52. 10.1111/j.1600-0765.2009.01204.x

Utsal, L., Tillmann, V., Zilmer, M., Maestu, J., Purge, P., Jurimae, J., Saar, M., Latt, E., Maasalu, K. & Jurimae, T. 2012. Elevated serum IL-6, IL-8, MCP-1, CRP, and IFN-gamma levels in 10- to 11-year-old boys with increased BMI. Horm Res Paediatr, 78, 31-9. 10.1159/000339831

58

Vaananen, H. K. & Zhao, H. 2002. Osteoclast Function. In: Bilezikian, J. P., Raisz, L. G. & Rodan, G. A. (eds.) Principles of Bone Biology. 2 ed. San Diego: Academic Press.

Verheul, M. K., Fearon, U., Trouw, L. A. & Veale, D. J. 2015. Biomarkers for rheumatoid and psoriatic arthritis. Clin Immunol, 161, 2-10. 10.1016/j.clim.2015.04.005

Villanova, F., Flutter, B., Tosi, I., Grys, K., Sreeneebus, H., Perera, G. K., Chapman, A., Smith, C. H., Di Meglio, P. & Nestle, F. O. 2014. Characterization of innate lymphoid cells in human skin and blood demonstrates increase of NKp44+ ILC3 in psoriasis. J Invest Dermatol, 134, 984-991. 10.1038/jid.2013.477

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. & 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. 10.1002/(SICI)1097-4652(200005)183:2<196::AID-JCP6>3.0.CO;2-8

Williams, S. R., Jiang, Y., Cochran, D., Dorsam, G. & Graves, D. T. 1992. Regulated expression of monocyte chemoattractant protein-1 in normal human osteoblastic cells. Am J Physiol, 263, C194-9.

Wintges, K., Beil, F. T., Albers, J., Jeschke, A., Schweizer, M., Claass, B., Tiegs, G., Amling, M. & Schinke, T. 2013. Impaired bone formation and increased osteoclastogenesis in mice lacking chemokine (C-C motif) ligand 5 (Ccl5). J Bone Miner Res, 28, 2070-80. 10.1002/jbmr.1937

Woo, C. H., Jeong, D. T., Yoon, S. B., Kim, K. S., Chung, I. Y., Saeki, T. & Kim, J. H. 2002. Eotaxin induces migration of RBL-2H3 mast cells via a Rac-ERK-dependent pathway. Biochem Biophys Res Commun, 298, 392-7.

Wright, L. M., Maloney, W., Yu, X., Kindle, L., Collin-Osdoby, P. & Osdoby, P. 2005. Stromal cell-derived factor-1 binding to its chemokine receptor CXCR4 on precursor cells promotes the chemotactic recruitment, development and survival of human osteoclasts. Bone, 36, 840-53. 10.1016/j.bone.2005.01.021

Xiong, J., Piemontese, M., Onal, M., Campbell, J., Goellner, J. J., Dusevich, V., Bonewald, L., Manolagas, S. C. & O'brien, C. A. 2015. Osteocytes, not Osteoblasts or Lining Cells, are the Main Source of the RANKL Required for Osteoclast Formation in Remodeling Bone. PLoS One, 10, e0138189. 10.1371/journal.pone.0138189

Yadav, A., Saini, V. & Arora, S. 2010. MCP-1: chemoattractant with a role beyond immunity: a review. Clin Chim Acta, 411, 1570-9. 10.1016/j.cca.2010.07.006

59

Yano, S., Mentaverri, R., Kanuparthi, D., Bandyopadhyay, S., Rivera, A., Brown, E. M. & Chattopadhyay, N. 2005. Functional expression of beta-chemokine receptors in osteoblasts: role of regulated upon activation, normal T cell expressed and secreted (RANTES) in osteoblasts and regulation of its secretion by osteoblasts and osteoclasts. Endocrinology, 146, 2324-35. 10.1210/en.2005-0065

Yoshida, H., Hayashi, S. I., Kunisada, T., Ogawa, M., Nishikawa, S., Okamura, H., Sudo, T., Shultz, L. D. & Nishikawa, S. I. 1990. The Murine Mutation Osteopetrosis Is in the Coding Region of the Macrophage Colony Stimulating Factor Gene. Nature, 345, 442-444. Doi 10.1038/345442a0

Yu, X., Antoniades, H. N. & Graves, D. T. 1993. Expression of monocyte chemoattractant protein 1 in human inflamed gingival tissues. Infect Immun, 61, 4622-8.

Yucel-Lindberg, T. & Bage, T. 2013. Inflammatory mediators in the pathogenesis of periodontitis. Expert Rev Mol Med, 15, e7. 10.1017/erm.2013.8

Zlotnik, A. & Yoshie, O. 2000. Chemokines: a new classification system and their role in immunity. Immunity, 12, 121-7.