The Role of Transglutaminases in Osteoclastogenesis Huifang Sun

The role of transglutaminases in osteoclastogenesis

Huifang Sun

Faculty of Dentistry McGill University, Montreal, Canada November 2019

A thesis submitted to McGill University as partial fulfillment of the requirements of the degree of Doctor of Philosophy in Craniofacial Health Sciences

CONTENTS

Abstract I Résumé II Acknowledgements IV Contributions of authors VI List of figures VII Abbreviations VIII

Chapter 1 - Introduction 1 1.1 Bone remodeling 1 1.1.1 Osteoblasts 1 1.1.2 Osteocytes 3 1.1.3 Bone lining cells 3 1.1.4 Osteoclasts 3 1.2 Osteoclast precursor commitment 5 1.2.1 Hierarchical model of hematopoiesis 5 1.2.2 Monocytes 5 1.2.3 Macrophages 5 1.3 Osteoclast differentiation 6 1.3.1 M-CSF signaling 6 1.3.2 RANKL signaling 7 1.3.3 Co-stimulatory signaling and calcium signaling 10 1.4 Osteoclast multinucleation 11 1.5 Osteoclast cytoskeletal organization 13 1.5.1 Podosome, podosome belt and sealing zone 13 1.5.2 Ruffled border and resorption 15 1.6 Cell culture models of osteoclastogenesis 16 1.7 Transglutaminases (TGs) 19 1.7.1 Structures and conformations of TGs 19 1.7.2 Enzymatic functions of TGs 22 1.8 Regulation of TG activity 25 1.9 TG-specific substrate peptides 26 1.10 TG inhibitors 26 1.11 TG family members 30

1.11.1 TG2 30 1.11.2 FXIII-A 31 1.11.3 TG1 32 1.11.4 Other TGs 33 1.12 TGs in monocytes and macrophages 34 1.13 TGs in bone remodelling 34 1.14 Hypothesis, rationale and objectives 35

Chapter 2 - Transglutaminases in monocytes and macrophages 37 2.1 Abstract 38 2.2 Introduction 38 2.3 Brief overview of TG family members 39 2.4 Development and classification of macrophages 40 2.5 TGs in monocyte adhesion and extravasation 41 2.6 TGs in macrophages 42 2.7 Conclusions 50 2.8 Acknowledgements 52

Chapter 3 - Transglutaminase activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics 53 3.1 Abstract 54 3.2 Introduction 55 3.3 Materials and methods 56 3.4 Results 61 3.5 Discussion 65 3.6 Acknowledgements 68 3.7 Figures 70

Chapter 4 - Assessment of expression and specific activities of transglutaminases TG1, TG2 and FXIII-A during osteoclastogenesis 88 4.1 Abstract 89 4.2 Introduction 90 4.3 Materials and methods 92 4.4 Results 95 4.5 Discussion 100 4.6 Acknowledgements 103 4.7 Figures 104

Chapter 5 - Conclusions, discussion and future directions 124 5.1 Original contributions 124 5.2 Conclusions and discussion 124 5.3 Future work 126

References 128

ABSTRACT

Osteoclasts are large multinucleated bone-resorbing cells derived from monocyte/macrophage lineage. Osteoclastogenesis is driven by two essential cytokines, macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor-κB ligand (RANKL). Increased osteoclast activity is responsible for bone destruction in diseases such as osteoporosis. Transglutaminases (TGs) are a family of Ca2+ and thiol-dependent acyl transferases that catalyze covalent cross-linking, i.e., isopeptide formation between a protein- or peptide-bound glutamine residue and a protein-bound lysine residue or a primary amine. TG family consists of eight catalytically active enzymes, TG1-7 and Factor XIII-A (FXIII-A) which share 100% homology of their active site residues yet show high specificity for their substrates. Abundant information exists on the involvement of TGs in monocyte and macrophage function, and our recent studies have shown that three TGs, TG1, TG2, and FXIII-A, are present in macrophages and osteoclasts. We and others have also demonstrated that TG2 and FXIII-A are involved in osteoclastogenesis in vivo and in vitro, however their mechanisms of action have remained unknown. This thesis project aimed at understanding the role of TGs in osteoclastogenesis. I report that TG-activity inhibitors block differentiation, migration, and fusion of pre-osteoclasts as well as disrupt podosome belt formation and resorption. In addition, TG inhibitor NC9, increased RhoA levels and its inhibitory effect on podosome belt formation and osteoclastogenesis was completely reversed with a Rho-family inhibitor exoenzyme C3, suggesting that TG activity regulates actin dynamics in osteoclasts. This thesis project also aimed at understanding the separate functions of the three TGs and assessing the total and specific activities of TG1, TG2, and FXIII-A during osteoclastogenesis in a day-by-day manner using 5-(biotinamido)pentylamine and biotinylated “Hitomi peptides”. I report that TGs activities are highest in the differentiation and early fusion phases and then decrease dramatically. TGs activities were upregulated by M-CSF and downregulated by addition of RANKL. FXIII-A was dramatically downregulated by RANKL, suggesting its involvement in M-CSF-mediated precursor commitment phase. TG1 and TG2 proteins were present throughout osteoclastogenesis, suggesting that they may have functions in both differentiation and fusion. In summary, this thesis shows that TG activity is required for osteoclastogenesis and regulates this process via affecting actin dynamics. Expression and activities of TGs are regulated by M-CSF and RANKL and the three TGs likely exert distinct functions at different stages of osteoclastogenesis.

RÉSUMÉ

Les ostéoclastes sont de grosses cellules multinuclées résorbant les os qui sont dérivées de la lignée cellulaire monocyte/macrophage. L'ostéoclastogenèse est dirigée par deux cytokines essentielles: le facteur de stimulation des colonies de macrophages (M-CSF) et l'activateur du récepteur du ligand du facteur nucléaire-κB (RANKL). L'activité accrue des ostéoclastes est responsable de la destruction des os dans des maladies telle que l'ostéoporose, par example. Les transglutaminases (TGs) sont une famille d'acyltransférases dépendant de groupe thiol et de Ca2+ qui catalysent la fixation covalente (c'est-à-dire la formation d’une isopeptide lors d’une réaction entre un résidu de glutamine d’une protéine ou d’un peptide et un résidu de lysine d’une protéine ou une amine primaire). La famille TG comprend huit enzymes catalytiquement actives, la TG1-7 et le facteur XIII-A (FXIII-A), qui partage 100% d'homologie des résidus de leur site actif, tout en montrant une haute spécificité pour leurs substrats. Il existe de nombreuses informations sur l'action des TGs dans la fonction des monocytes et des macrophages et nos études récentes ont montré que trois TGs, à savoir la TG1, la TG2 et le FXIII-A, sont présentes dans les macrophages et les ostéoclastes. Nous, ainsi que d’autres groupes de recherches, nous avons également démontré que la TG2 et le FXIII-A sont impliqués dans l'ostéoclastogenèse in vivo et in vitro, mais leurs mécanismes d'action sont restés inconnus. Ce projet de thèse vise à comprendre le rôle des TGs dans l'ostéoclastogenèse. Je démontre que les inhibiteurs de la TG bloquent la différenciation, la migration et la fusion des cellules préostéoclastes, ainsi que la perturbation de la formation et de la résorption de la ceinture des podosomes. De plus, l’inhibiteur NC9 de la TG augmente l’expression de RhoA. Lorsque RhoA est inhibée avec l’exoenzyme C3, les effets de l’inhibiteur NC9 sur la formation de la ceinture des podosomes et de l'ostéoclastogenèse ont été complètement inversés. Ceci suggère que l'activité de la TG régule la dynamique de l'actine dans les ostéoclastes. Ce projet de thèse vise également à comprendre les fonctions distinctes des trois TGs mentionnées en haut (TG1, TG2 et FXIII-A) et à évaluer de manière quotidienne leur activité totale et leurs activités spécifiques au cours de l'ostéoclastogenèse. Les activités ont été mesurées au moyen de la 5-(biotinamido)pentylamine ou avec des «peptides Hitomi» biotinylés. Je démontre que les activités des TGs sont plus élevées dans les phases de différenciation et de fusion précoce et puis elles diminuent fortement. Les activités des TGs ont été régulées positivement par le M-CSF et régulées négativement par l'ajout de RANKL. L’expression du FXIII-A a été considérablement réduite par RANKL, suggérant sa participation dans la phase de

II l’engagement des précurseurs médiée par M-CSF. La TG1 et la TG2 sont restés présentes pendant l'ostéoclastogenèse, suggérant qu’elles peuvent avoir de fonctions aux stades de différenciation et de fusion. En résumé, cette thèse montre que l’activité de la TG est nécessaire à l’ostéoclastogenèse et régule ce processus en affectant la dynamique de l’actine. Les expressions et les activités des TGs sont régulées par M-CSF et RANKL et la TG1, la TG2 et le FXIII-A exercent tous probablement de fonctions distinctes à différents stades de l'ostéoclastogenèse.

III

ACKNOWLEDGEMENTS

I would like to express my deepest and most sincere gratitude to my supervisor, Dr. Mari T. Kaartinen, who has given me a great deal of support and encouragement throughout my PhD. Firstly, she offered me the invaluable opportunity to study at the prestigious McGill University and set up such an interesting and challenging project for me. Along these years, she is always there to help me with her expert guidance, patience and trust. In addition, I am indebted to her for giving me opportunities to attend several international conferences to present my work and meet leading researchers in the field. I am especially appreciative of the time and energy she has devoted to reading and correcting my presentations, posters, manuscripts and this thesis. Her scientific acumen has shaped my academic aptitudes and will continue to nourish my research career in the future. I would also like to thank my advisory committee members, Dr. Svetlana V. Komarova, Dr. Dieter P. Reinhardt and Dr. Pierre Moffatt for their unconditional support, time and mentorship. I have learned immensely from them, both personally and professionally. Thanks to all my past and current colleagues who have been part of this journey, together we have discovered the difficulties and rewards of science. Also many thanks to Cui Cui and Aisha Mousa for all the experimental techniques they have taught me. I would like to highlight my endless appreciation to Dr. Marc D. McKee, Dr. Simon D. Tran, Dr. Khanh Huy Bui, and the Alan Edwards Centre for Research on Pain for sharing their equipment and reagents. Without them, so much of the work herein would not have been possible. My appreciation also goes to all my friends for the cordial support I have received and the unforgettable time we have spent together. I feel truly blessed to have them around, making my life bright and exciting. My deepest appreciation goes to my beloved family for always being there for me, always believing in me, and for giving me everything that a daughter, sister and wife would ever dare to wish and ask for. And a special thanks goes to Peng Li, my best friend and husband, for giving me enormous freedom to explore the world. It is his infinite unconditional love that has magnified the best of me and kept me going. I can never repay him for all the sacrifices he made all these years. I am sure we will spend more time together from now on and I look forward to the new adventures in our life. Ultimately, I would like to extend my thanks to the Faculty of Dentistry. Everyone has been so helpful, so approachable and so intelligent. I do not have enough space to list the names of all

IV outstanding individuals I have encountered along the way. The six-year studies at McGill have been the most rewarding experience in my life and I will truly miss it.

CONTRIBUTIONS OF AUTHORS

This thesis includes published papers and an unpublished manuscript of which the candidate is first author. Contributions of all authors are listed below:

1. Sun H and Kaartinen MT (2018). Transglutaminases in monocytes and macrophages. Med Sci. 6(4). pii: E115.

 H Sun performed the literature search and wrote the manuscript.  MT Kaartinen edited the manuscript.

2. Sun H and Kaartinen MT (2018). Transglutaminase activity regulates differentiation, migration and fusion of osteoclasts via affecting RhoA activity and actin dynamics. J Cell Physiol. 233(9):7497-7513.

 H Sun designed and performed all experiments and data analysis and wrote the manuscript.  MT Kaartinen assisted in designing the experiments, supervised the study and edited the manuscript.

3. Sun H and Kaartinen MT. Assessment of expression and specific activities of transglutaminases TG1, TG2 and FXIII-A during osteoclastogenesis. Anal Biochem. In revision.

 H Sun designed and performed all experiments and data analysis and wrote the manuscript.  MT Kaartinen assisted in designing the experiments, supervised the study and edited the manuscript.

LIST OF FIGURES

Figure I Proposed model for bone remodeling. 2 Figure II Stages of osteoclastogenesis. 4 Figure III Major RANK signaling pathways in osteoclastogenesis. 12 Figure IV Scheme of podosome belt and sealing zone formation in cultured osteoclasts. 17 Figure V Bone resorption and ruffled border formation. 18 Figure VI Phylogenetic tree of transglutaminase family. 20 Figure VII Protein domains and conformational changes of transglutaminase. 21 Figure VIIIa Transglutaminases catalyze various post-translational reactions. 23 Figure VIIIb Transglutaminases catalyze various post-translational reactions. 24 Figure IX Sequences of transglutaminase-specific substrate peptides. 28 Figure X Structures and function groups of transglutaminase inhibitors. 29 Figure XI The role of transglutaminases in osteoclastogenesis at different stages. 127

VII

ABBREVIATIONS

TG Transglutaminase FXIII-A Factor XIII-A M-CSF Macrophage-colony stimulating factor RANKL Receptor activator of nuclear factor-κB ligand OPG Osteoprotegerin GM-CSF Granulocyte macrophage-colony stimulating factor BMU Basic multicellular unit HSC Hematopoietic stem cell BMMSC Bone marrow mesenchymal stem cell COL I Type I collagen BSP Bone sialoprotein OPN Osteopontin FN Fibronectin TRAFs Tumor necrosis factor receptor associated factors NFATc1 Nuclear factor of activated T cell c1 TAK1 TGF-β-activated kinase 1 IKK Inhibitory κB kinase MITF Microphthalmia transcription factor AP-1 Activator protein-1 DC-STAMP Dendritic cell-specific transmembrane protein Atp6v0d2 d2 isoform of vacuolar ATPase V0 domain OC-STAMP Osteoclast stimulatory transmembrane protein OSCAR Osteoclast-associated receptor TREM-2 Triggering receptor expressed in myeloid cells-2 TRAP Tartrate-resistant acid phosphatase MMP-9 Matrix metalloproteinase-9 CTSK Cathepsin K ITAM Immunoreceptor tyrosine-based activation motif FcRγ Fc receptor common γ subunit DAP12 DNAX-activating protein 12 GEF Guanine nucleotide exchange factor

VIII

GAP GTPase-activating protein FSD Functional secretory domain NO Nitric oxide MDC Monodansylcadaverine bPA 5-(biotinamido)pentylamine PDI Protein disulfide isomerase MT1-MMP Membrane-type matrix metalloproteinase-1

Chapter 1 - Introduction

1.1 Bone remodeling Bone is a mineralized connective tissue and provides important functions, such as structural support of the body, protection for vital organs, attachment sites for muscles, calcium and phosphate storage, and harboring of the bone marrow [1]. Despite its rigidity and hardness, bone is a highly dynamic organ that constantly undergoes remodeling during life. Bone remodeling is a complex process in which old, microdamaged bone is resorbed by osteoclasts and new, mechanically stronger bone is formed by osteoblasts [2]. The remodeling process involves five phases: activation, resorption, reversal, formation and termination (Figure I) [3]. Normal bone remodeling depends on the highly coordinated activities of osteoblasts, osteoclasts, osteocytes, and bone lining cells which together form the anatomic and functional structure termed basic multicellular unit (BMU) [2].

1.1.1 Osteoblasts Osteoblasts originate from bone marrow mesenchymal stem cells (BMMSCs) and are responsible for bone matrix synthesis and the subsequent mineralization [4]. Osteoblasts are mononucleated cuboidal cells, lining up along the bone-forming surface. A mature osteoblast shows typical morphological features of protein synthesizing cells, such as a large nucleus, enlarged Golgi apparatus, and extensive endoplasmic reticulum (ER) [5]. Initially, osteoblasts secrete abundant type I collagen (COL I) and non-collagenous proteins such as osteocalcin, osteonectin, bone sialoprotein (BSP), osteopontin (OPN), and proteoglycans, which together form the osteoid [6]. Thereafter, this osteoid serves as a template for the subsequent deposition of mineral in the form of hydroxyapatite. Tissue-nonspecific alkaline phosphatase (TNAP), a membrane-bound marker of osteoblast differentiation, is necessary to promote mineralization by providing phosphate from potential phosphate-generating species, and by decreasing the concentration of pyrophosphate, a potent calcification inhibitor [7].

FIGURE I

HSC Bone marrow space MPP

LMPP

NMP

Monocyte BMMSC/Osteoprogenitor

BMU

Monocyte BMMSC/Osteoprogenitor Bone lining cell Osteoid

Osteoclast Osteoblast Osteocyte Bone

FIGURE I Proposed model for bone remodeling. The remodeling process involves five distinct stages: (1) activation of bone remodeling at a specific site; (2) predominant bone resorption phase; (3) transition signals generated to halt bone resorption and stimulate bone formation process; (4) predominant osteoblast differentiation and function (osteoid synthesis); and (5) mineralization of osteoid and completion of bone remodeling. HSC: hematopoietic stem cell; MPP: multipotent progenitor; LMPP: lymphoid-primed multipotent progenitor; NMP: neutrophil-monocyte progenitor; BMMSC: bone marrow mesenchymal stem cell; BMU: basic multicellular unit. Modified from [46].

2 1.1.2 Osteocytes As bone matrix deposition continues, osteoblasts become embedded in the osteoid and transform into osteocytes. Osteocytes are the most abundant cell type of the bone, accounting for 95% of all bone cells [8]. Mature osteocytes are stellate-shaped cells, residing in a lacunar space. They communicate with each other, as well as with osteoblasts and bone lining cells on the bone surface via numerous long and slender cell processes which transverse the bone matrix through small canals called canaliculi [9]. Osteocytes can modulate bone remodeling by regulating osteoclast differentiation via producing pro- and anti-osteoclastogenic cytokines (as discussed below), and regulating osteoblast activity via secreting sclerostin, an inhibitor of bone formation [10]. Osteocytes also act as mechanosensors to coordinate adaptive responses of the skeleton to mechanical loading. In addition, an emerging role of osteocytes as endocrine cells to regulate phosphate homeostasis has been discovered recently [11].

1.1.3 Bone lining cells At the end of the bone formation, osteoblasts can transform into inactive quiescent bone lining cells. They are flat spindle-shaped cells extending along the bone surfaces and can connect to each other as well as to osteocytes via gap junctions [1]. Bone lining cells can convert to active osteoblasts with mechanical or hormonal stimulation [12]. They can also initiate bone formation by both cleaning matrix left over by osteoclasts and by depositing thin collagen fibers [13]. In addition, bone lining cells also function as important gatekeepers of estrogen- controlled bone resorption [14].

1.1.4 Osteoclasts Osteoclasts, the multinucleated bone-resorbing cells, are derived from monocyte/macrophage lineage of the hematopoietic stem cells (HSCs). For osteoclasts to be formed, two key factors, namely macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor κB ligand (RANKL) are required [15]. Osteoclastogenesis is a complicated process involving several stages, which will be discussed below. They include osteoclast precursor commitment, differentiation, multinucleation, and resorption (Figure II). In addition, there is a dynamic cytoskeletal organization concomitant with the migration, multinucleation and resorption [16].

FIGURE II

HSC

M-CSF Osteoclast precursor commitment

Monocyte/macrophage

M-CSF RANKL Differentiation

TRAP+ pre-osteoclast

M-CSF RANKL Multinucleation

Multinucleated osteoclast

M-CSF RANKL Resorption

Active osteoclast

FIGURE II Stages of osteoclastogenesis. Osteoclasts are derived from monocyte/macrophage lineage of the hematopoietic stem cells (HSCs). For osteoclasts to be formed, two key factors, namely macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor κB ligand (RANKL) are required. Osteoclastogenesis is a complicated process involving several stages, including osteoclast precursor commitment, differentiation, multinucleation, and resorption.

4 1.2 Osteoclast precursor commitment 1.2.1 Hierarchical model of hematopoiesis All the hematopoietic lineage cells, including osteoclasts, arise from HSCs in the bone marrow. HSCs are multipotent and undergo self-renewal. HSCs initially yield the multipotent progenitors (MPPs) which have lost their self-renewal capacity but still retain their multipotency [17]. Further downstream, MPPs bifurcate into the erythroid-megakaryocyte primed multipotent progenitors (EMkMPPs), and the lymphoid-primed multipotent progenitors (LMPPs) [18]. The EMkMPPs possess combined erythroid/megakaryocyte and myeloid potential, and give rise to erythrocytes, megakaryocytes, eosinophils, and basophils or mast cells. LMPPs possess combined lymphoid and myeloid potential and bifurcate further into common lymphoid progenitors (CLPs) and neutrophil-monocyte progenitors (NMPs) [18]. CLPs differentiate into B cells, T cells, and natural killer cells (NK cells). NMPs differentiate into neutrophils and monocytes [18].

1.2.2 Monocytes Monocytes constitute 10% of total leukocytes in human blood and 4% in mouse blood, respectively [19,20]. In mouse, two discrete subsets of monocytes have been identified based on their differential expression of surface markers Ly6C, CCR2 and CX3CR1 [21]. The + high low Ly6C CCR2 CX3CR1 monocytes can traffic across the blood vessel wall and migrate into selected tissues and differentiation to macrophages at steady state. In pathological conditions, they are rapidly recruited into the damaged and inflamed tissues and give rise to cells with macrophage or dendritic cell features depending on the local cues. In addition, they can also − low high differentiate into the Ly6C CCR2 CX3CR1 monocytes, which constantly patrol and survey the endothelium by crawling along the luminal surface of small vessels [22]. Monocyte development depends on M-CSF, also known as CSF-1, and its receptor c-Fms, also known as M-CSFR, CSF1R, CD115 [19]. In mice deficient in M-CSF and c-Fms, the number of blood monocytes is significantly reduced [23-25]. In addition, Src family kinase signaling is required to mediate monocyte differentiation induced by M-CSF [26].

1.2.3 Macrophages Macrophages are immune cells present in almost all tissues of the body with distinct tissue- specific phenotypes and functions, playing a central role in inflammation and tissue homeostasis via phagocytosis and orchestrating immune responses [27]. Macrophages are

5 positive for F4/80 in mice [28] and can be classified based on their anatomical locations and functional phenotypes, such as splenic macrophages, peritoneal macrophages, alveolar and interstitial macrophages (lung), Kupffer cells (liver), osteoclasts (bone), and microglia (central nervous system, CNS) [29]. Macrophages have a plasticity and respond to various stimuli in the surrounding environment by mounting a specific activation phenotype, i.e., classically activated (M1) and alternatively activated (M2) macrophages [30]. M1 is activated by lipopolysaccharide (LPS)/interferon gamma (IFN-γ) and exhibits pro-inflammatory features, whereas M2 results from IL-4/IL-13 stimulation and displays anti-inflammatory properties [30]. Notably, granulocyte macrophage-colony stimulating factor (GM-CSF) and M-CSF can mediate M1 and M2 polarization of macrophages, respectively [31].

1.3 Osteoclast differentiation 1.3.1 M-CSF signaling M-CSF is expressed by various cells, such as vascular smooth muscle cells and endothelial cells [32,33], hepatocytes [34], fibroblasts [35], T cells [36], BMMSCs [37], osteoblasts [38], and osteocytes [39]. Apparently, BMMSCs, osteoblasts, and osteocytes are the major sources of M-CSF for osteoclast formation. M-CSF exerts its functions via binding and activating its receptor c-Fms, a type III receptor tyrosine kinase, which contains an Ig-like extracellular ligand-binding domain, a single transmembrane domain, an intracellular juxtamembrane domain, a split tyrosine kinase domain, and a C-terminal tail [40]. M-CSF-induced dimerization of c-Fms monomers results in the activation of c-Fms and enhanced tyrosine kinase activity, leading to the trans-phosphorylation of specific tyrosine residues in the cytoplasmic domain [41], i.e., Y544 [42], Y559, Y697, Y706, Y721, Y807 [43], Y921 [44], and Y974 [45]. These phosphorylated tyrosine residues then act as docking sites for downstream signaling molecules. Studies showed that six of the eight tyrosine residues, i.e., Y544, Y559, Y697, Y721, Y807, and Y921, played a role in promoting the proliferation of macrophages, known as osteoclast precursors [46]. Phospho-Y544 in juxtamembrane domain was shown to bind an unidentified 55kd protein and was required for macrophage proliferation [47]. No adaptor protein has been reported to bind phospho-Y807, which is located in the activation loop. Therefore, it is proposed to participate in the structural activation of c-Fms rather than inducing signaling pathways [47]. Phospho-Y559 is required for the full activation of c-Fms via interacting with c-Src [48]. Then the phospho-Y559/c-Src complex recruits the phosphatidylinositol 3-kinase (PI3K) which in turn activates the Akt pathway [49]. This

6 complex also recruits c-Cbl, leading to the ubiquitination of c-Fms [50]. The c-Cbl-induced ubiquitination promotes the tyrosine phosphorylation and activation of c-Fms via a conformational change in the kinase domain. In contrast to phospho-Y559, phospho-Y721 can interact with PI3K directly to activate Akt [49,51]. In addition, phospho-Y697 and phospho- Y921 recruits Grb2 [44], leading to the activation of ERK pathway. Taken together, phospho- Y559, phospho-Y721, phospho-Y697, and phospho-Y921 promote macrophage proliferation via the ERK and PI3K/Akt pathways. It has been demonstrated that M-CSF is pivotal for the survival of macrophages as well [52]. Although the specific tyrosine residues are not identified yet, it is likely that phospho-Y559, phospho-Y721, phospho-Y697, and phospho-Y921 are involved in this event due to the essential role of ERK and PI3K/Akt pathways in prolonging the survival of macrophages [53- 55]. In addition to its role in promoting proliferation and survival of macrophages, M-CSF is also necessary for osteoclast formation by priming the macrophages to respond to RANKL by inducing its receptor RANK in a Fos-dependent manner [56]. Consistently, studies showed that mice in which csf1r gene (coding c-Fms) was inactivated, developed a severe osteopetrotic phenotype due to a complete absence of osteoclasts, similar to the phenotypes of animal models in which loss-of-function mutations in csf1 gene (coding M-CSF) naturally occurred (the op/op mice or the tl/tl rats) [23,57-59]. Interestingly, despite their overall similarity, the phenotypes of the csf1r-/- mice are more severe than those of the op/op mice, leading to the discovery of a new cytokine, IL-34. IL-34, sharing no sequence homology with M-CSF, is another ligand of c-Fms [60]. Although IL-34 is much less potent than M-CSF due to its lower affinity for c-Fms, it promotes the proliferation of macrophages via activation of ERK 1/2. More importantly, IL- 34 and RANKL are sufficient to induce osteoclast formation in the absence of M-CSF. Additionally, transgenically expressed IL-34 in an M-CSF-specific manner is able to rescue macrophage and osteoclast defects of op/op mice [61]. However, given the different spatiotemporal expression of IL-34 and M-CSF, and a complete lack of osteoclasts in op/op mice and tl/tl rats, it is likely that IL-34 is not important for normal bone remodeling. A recent study demonstrated that IL-34 played a role in osteoclast formation in rheumatoid arthritis [62].

1.3.2 RANKL signaling RANKL, another key osteoclastogenic factor, is a type II transmembrane protein of the tumor necrosis factor (TNF) superfamily. It contains a short N-terminal intracellular tail and a C-

7 terminal extracellular receptor-binding domain [63]. Soluble RANKL is generated either by proteolytic processing of the membrane-bound RANKL by matrix metalloproteinases or by alternative splicing [64,65]. RANKL, encoded by Tnfsf11, self-associates as a homotrimer. RANKL with a mutation at codon 278 (RANKLG278R) is unable to bind and activate its receptor RANK since the monomers fail to assemble into homotrimers [66]. RANKL is expressed in a variety of cell types, such as BMMSCs [67,68], osteoblasts [67,69], osteocytes [70], bone lining cells [14], T cells [71], B cells [72], and adipocytes [73]. Studies demonstrated that hypertrophic chondrocytes and osteoblasts were the major source of RANKL in bone metabolism in the growth period of young mice [74], whereas osteocytes were the main source of RANKL in the adult mice [70]. RANKL exerts its effects via activating its receptor RANK, which is a type I transmembrane protein of the TNF receptor (TNFR) superfamily. RANK is encoded by Tnfrsf11a and it consists of four tandem cysteine-rich pseudo-repeat domains (CRDs) among which the CRD2 and CRD3 can directly bind to RANKL [75]. Upon binding to RANKL, RANK trimerizes and activates the downstream signaling pathways via adaptor molecules [76]. Osteoprotegerin (OPG) belongs to the TNFR superfamily as well. OPG is encoded by Tnfrsf11b and is expressed by BMMSCs [77], osteoblasts [78], osteocytes [79], fibroblasts [80], endothelial cells [81,82], lymphoid cells [83], and smooth muscle cells [83]. OPG is a soluble decoy receptor for RANKL and blocks RANKL functions through sequestering RANKL and preventing its binding to RANK [84]. Therefore, the transduction of the RANKL signal is determined by the ratio of RANKL to OPG and RANK expression on osteoclast precursors [85]. Numerous studies on genetically modified mice have demonstrated the significance of RANKL/RANK/OPG signaling in bone metabolism. Tnfsf11-/- mice, Tnfrsf11a-/- mice and mice overexpressing Tnfrsf11b gene developed severe osteopetrosis [86-90]. Consistently, Tnfrsf11b-/- mice exhibited osteoporosis [91,92]. In addition, mutations in RANKL, RANK and OPG have been identified in patients with bone disorders. Autosomal recessive osteopetrosis results from mutations of RANKL affecting its trimerization. Familial expansile osteolysis, expansile skeletal hyperphosphatasia, and early-onset Paget’s disease are rare bone diseases caused by constitutive activation of RANK. Juvenile Paget’s disease, which is another type of Paget’s disease, arises from mutations in OPG [93]. Upon binding to RANKL, RANK adopts a trimeric conformation and recruits TNF receptor associated factors (TRAFs) to activate various signaling pathways to mediate

8 osteoclastogenesis [76]. Five TRAF proteins (TRAF 1, 2, 3, 5 and 6) are able to interact with RANK. Among TRAFs, TRAF6 is essential for osteoclastogenesis, as evidenced by the severe osteopetrosis developed in Traf6-/- mice [94,95]. TRAF6 can form a complex with TGF-β- activated kinase 1 (TAK1) and TAK1-binding protein 2 (TAB2) [96]. When the TRAF6- TAB2-TAK1 complex is formed, TAK1 activates NF-кB and MAPKs (p38 and JNK) pathways which subsequently induce the expression of nuclear factor of activated T cell c1 (NFATc1), which is a master regulator of osteoclastogenesis (Figure III) [76]. Interestingly, a new RANK motif (IVVY535-538) has been identified, which is involved in the early commitment of macrophages to the osteoclast lineage [97]. This motif also plays an essential role in TNF-α- and IL-1-mediated osteoclastogenesis [98,99]. NF-κB family consists of five proteins, namely RelA, RelB, c-Rel, NF-κB1 (p50 and its precursor p105), and NF-κB2 (p52 and its precursor p100) [100]. Both p50 and p52 lack the transcriptional activation domains, therefore, they need to heterodimerize with one of the three Rel proteins for activation function [101]. In resting cells, NF-κB dimer remains inactive in the cytosol but rapidly enter the nucleus upon activation by two pathways [102]. In the classic pathway, inhibitory κB kinase (IKK) complex, whose critical components are IKKβ and IKKγ [100], phosphorylates the inhibitors of NF-κB (IκBs) and targets them for degradation by the proteosome, leading to the nuclear translocation of RelA/p50. In the alternative pathway, NF- κB-inducing kinase (NIK) activates the homodimer IKKα [103], which cleaves p100 to generate p52, leading to the nuclear translocation of RelB/p52. As mentioned above, TAK1 activates NF-κB via activating the IKK [96]. Activated NF-κB then induces the activation of c-Fos [104]. Studies have shown that TAK1-deficient mice [105], IKKβ-deficient mice [106], p50/p52 double knockout mice [107], and c-Fos-deficient mice [108] all develop osteopetrosis, indicating the significance of NF-κB signaling in osteoclastogenesis. TAK1 also acts as a MAPK kinase kinase (MAPKKK) to phosphorylate the MAPK kinases (MAPKKs) which subsequently phosphorylate the MAPKs, including p38 and c-Jun N- terminal kinase (JNK). MKK6, an MAPKK, plays a role in p38 phosphorylation during osteoclastogenesis [109]. p38α ablation in monocytes and macrophages causes a minor increase in bone mass in young mice [110]. MKK7 is another MAPKK which phosphorylates JNK. Overexpression of a dominant negative form of MKK7 inhibits RANKL-induced JNK phosphorylation, leading to the suppressed phosphorylation of c-Jun and osteoclast formation [111]. JNK1-deficient bone marrow cells fail to differentiate into osteoclasts due to the impaired phosphorylation of c-Jun [112]. Consistently, c-Jun-deficient bone marrow cells show

9 a decreased osteoclast differentiation [113]. JunB, another member of the Jun proteins, is also involved in osteoclastogenesis, since JunB-deficient mice exhibit an osteopetrotic phenotype [114]. Downstream of NF-κB and MAPK signaling, c-Fos and Jun proteins (c-Jun and JunB) are activated and subsequently dimerize to form activator protein-1 (AP-1). Then NF-κB, AP-1, NFATc2, activating transcription factor 4 (ATF4) and Jun dimerization protein 2 (Jdp2) are collectively recruited to the promoter region of the Nfatc1 gene, triggering expression of NFATc1, which is a master regulator of osteoclastogenesis [115-118]. NFATc1-deficient mice exhibit osteopetrosis owing to impaired osteoclastogenesis [119]. NFATc1-deficient cells fail to differentiate into osteoclasts, and ectopic expression of NFATc1 in osteoclast precursors induce osteoclastogenesis even in the absence of RANKL [120]. Ultimately, NFATc1 together with other transcription factors such as PU.1, microphthalmia transcription factor (MITF), and AP-1 drives various osteoclastogenic genes including Dcstamp (dendritic cell-specific transmembrane protein: DC-STAMP), Atp6v0d2 (d2 isoform of vacuolar ATPase V0 domain), Ocstamp (osteoclast stimulatory transmembrane protein: OC-STAMP), Oscar (osteoclast- associated receptor: OSCAR), Itgb3 (integrin β3), Acp5 (tartrate-resistant acid phosphatase: TRAP), Mmp9 (matrix metalloproteinase-9: MMP-9), Calcr (calcitonin receptor), and Ctsk (cathepsin K: CTSK) [120-125].

1.3.3 Co-stimulatory signaling and calcium signaling After the initial induction of NFATc1 by NF-κB, AP-1, NFATc2, ATF4 and Jdp2, calcium/calmodulin signaling is required for robust amplification and its translocation to the nucleus. However, RANK does not directly initiate calcium signaling, therefore, the costimulatory signaling for RANK is necessary for the full activation of NFATc1 via calcium/calmodulin signaling pathways [126]. In addition to RANK, immunoglobulin-like receptors are expressed on osteoclasts, such as OSCAR and triggering receptor expressed in myeloid cells-2 (TREM-2) [127]. These receptors contain an extracellular region which possesses an immunoglobulin-like domain, and an intracellular region which can interact with adaptor proteins containing the immunoreceptor tyrosine-based activation motif (ITAM) such as Fc receptor common γ subunit (FcRγ) and DNAX-activating protein 12 (DAP12) [128]. After the ITAM of FcRγ and DAP12 is phosphorylated, possibly induced by RANKL stimulation [126], spleen tyrosine kinase (Syk) is activated, leading to the formation of a complex, containing Bruton’s tyrosine kinase (Btk),

Tec, B cell linker protein (BLNK), and Src homology 2 domain-containing leukocyte protein of 76 kDa (SLP-76) [129]. Subsequently, this complex activates of PLCγ, mediating the hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) in the cell membrane, generating inositol 1,4,5-triphosphate (IP3) [129]. IP3 binds to its receptor IP3R on the surface of the ER, leading to the Ca2+ releasing into the cytoplasm. Sarco/endoplasmic reticulum Ca2+ ATPase 2 (SERCA2) then reuptakes the released Ca2+ into ER, causing a repetitive fluctuation of Ca2+ in the cytoplasm, i.e. calcium oscillation [130]. Calcium binds to calmodulin, which in turn activates the calmodulin-dependent phosphatase calcineurin, leading to the dephosphorylation of NFATc1 and its translocation into the nucleus [125].

1.4 Osteoclast multinucleation Multinucleation of osteoclasts, by the cell-cell fusion, is an essential step in osteoclastogenesis, as mononucleated osteoclasts cannot resorb bone efficiently [131]. Cell-cell fusion is achieved by the remodeling of phospholipids, components of the lipid bilayer of cell membrane. Phosphatidylethanolamine (PE) normally reside on the inner leaflet. However, in fusing osteoclasts, PE is greatly increased on outer leaflet of the filopodia, which are protruded “neurite-like” structures, forming contacts with their neighboring cells. Immobilization of the cell surface PE blocked osteoclast fusion [132]. In addition, there are three main molecules involved in the fusion of osteoclasts, i.e., DC-STAMP, OC-STAMP, and Atp6v0d2. DC- STAMP is the master regulator of osteoclast fusion [133] and its deficiency results in TRAP+ mononuclear cells both in vitro and in vivo due to complete blocked cell-cell fusion [134,135]. Furthermore, DC-STAMP heterogeneity determines the fusion potential of osteoclast precursors, i.e., DC-STAMP(lo) cells are the master fusogens, whereas DC-STAMP(hi) cells can only act as mononuclear donors [136]. Similarly, cells deficient in OC-STAMP also give rise to TRAP+ mononuclear cells both in vitro and in vivo [137-139]. Although DC-STAMP and OC-STAMP share many common features, they are not interchangeable since the deficiency of one protein cannot be complemented by overexpression of the other one [133]. Atp6v0d2 is a subunit of V-ATPase and its deficiency results in fewer multinucleated osteoclasts both in vitro and in vivo [140].

FIGURE III

RANKL

RANK

TRAF6 TAK1 TAB2

NF-кB JNK p38

c-Fos c-Jun JunB

Cytosol

Nucleus NF-кB AP-1 NFATc2 NFATc1 ATF4 Jdp2 PU.1

MITF

Osteoclastogenic genes

FIGURE III Major RANK signaling pathways in osteoclastogenesis. Upon binding to RANKL, RANK recruits TNF receptor associated factor 6 (TRAF6) which can form a complex with TAK1- binding protein 2 (TAB2) and TGF-β-activated kinase 1 (TAK1). TAK1 in turn activates NF-кB and MAPKs (p38 and JNK) pathways. Activated NF-κB then induces the activation of c-Fos, and JNK induces the activation of c-Jun and JunB. Subsequently, c-Fos and Jun proteins (c-Jun and JunB) form activator protein-1 (AP-1), and then NF-κB, AP-1, NFATc2, ATF4 and Jdp2 collectively trigger the expression of nuclear factor of activated T cell c1 (NFATc1), which is a master regulator of osteoclastogenesis. Ultimately, NFATc1 together with other transcription factors such as PU.1 and microphthalmia transcription factor (MITF) drives various osteoclastogenic genes. 12 1.5 Osteoclast cytoskeletal organization Cytoskeletal organization is not isolated stage and it occurs concomitantly with the migration, multinucleation, and resorption of osteoclasts.

1.5.1 Podosome, podosome belt and sealing zone Cells from monocytic lineage, such as macrophages and osteoclasts, develop podosomes to adhere and migrate with [141]. Podosome exert forces where tension is needed, and they tend to form in expanding regions and disappear from retracting regions [142]. Podosomes are highly dynamic dot-like structures, containing an F-actin-rich core and a surrounding F-actin cloud. The core consists of densely packed actin filaments extending perpendicularly to the plasma membrane, while the cloud consists of less dense actin filaments radiating from the core like a dome [143]. Adhesion of the actin filaments within the core to the substrate is mediated by the hyaluronate receptor CD44, whereas adhesion of the actin filaments within the cloud to the substrate is via integrin αvβ3 and integrin-associated proteins such as talin, kindlin, and vinculin [144]. The core also contains various actin-associated proteins such as cortactin and gelsolin, as well as actin-regulatory proteins such as Wiskott-Aldrich syndrome protein (WASP), neural WASP (N-WASP), and Arp2/3 [145]. The dynamic podosomes have a half- life of 2 to 12 min [146] and can organize into different superstructures on different substrates (Figure IV) [147]. On non-mineralized substrate, podosomes initially pattern into clusters. As the cell matures, podosomes are arranged into short-lived internal rings which expand to the cell periphery to form a stable podosome belt where the podosomes become denser and tightly interconnected [148]. Notably, podosome rings generate forces that drive osteoclast migration via extending the cellular membrane and push the substrate outward [142]. In addition, the microtubules promote the stabilization of the podosome belt but are not required for the formation of podosome clusters and rings [148]. On the other hand, on mineralized substrate, podosomes initially pattern into the actin patches which are uncharacterized actin clusters, and finally develop into a thickened, stable annular structure referred to as sealing zone where the podosomes become more densely packed and more interconnected [149]. Sealing zone and podosome belt only differ in the density and degree of interconnectivity of their building blocks, i.e., podosomes [143]. In addition, osteoclasts are spread out on non-mineralized substrate, whereas they round up on mineralized substrate, acquiring an apico-basal polarized phenotype and are able to resorb matrix [150]. Rho subfamily of small GTPases are known as critical regulators of the actin cytoskeleton,

13 including the best characterized RhoA, Rac1, and Cdc42 [151]. Rho GTPases are converted from the inactive form (GDP-bound) to the active form (GTP-bound) by a guanine nucleotide exchange factor (GEF) [152]. The active Rho GTPases become inactive via their own GTPase activity, which can be promoted by GTPase-activating protein (GAP) [152]. Both c-Fms and RANK are able to increase the activation of Rac via phosphorylation of Vav3 (a Rac GEF) in a c-Src-dependent manner in osteoclasts [153,154]. To study the role of RhoA in osteoclasts, C3 exoenzyme, an inhibitor of RhoA, B, and C, are widely used [155]. Studies show both Rho inhibition with C3 treatment and RhoA activation by expressing constitutively active RhoA in avian osteoclast-like cells result in podosome dissolution [156,157], suggesting that RhoA activity should be tightly regulated in a narrow optimal window to promote the podosome formation. In addition, RhoA also regulates podosome patterning [158]. Studies show that high Rho activity is required for sealing zone formation since Rho inhibition with C3 disrupts the sealing zone and triggers the podosome belt formation in osteoclasts seeded on bone. Concomitantly, Rho inhibition induces a transition from a polarized and rounded up morphology to one that is flat and spread out [150]. However, microinjection of active RhoA is not sufficient to induce sealing zone formation in osteoclasts seeded on glass [159]. On the other hand, Rho inhibition with C3 accelerated podosome belt formation in osteoclasts seeded on glass [159], indicating that lower Rho activity in cells is required for podosome belt formation. Myosin IXB is a Rho GAP and its suppression leads to an increase in the activity of Rho in osteoclasts. Myosin IXB suppression inhibits the formation of podosome belts in osteoclasts on glass. Interestingly, osteoclasts with suppressed Myosin IXB form sealing zones normally on bone [160]. Rac1/Rac2 double knockout osteoclasts do not form sealing zones on bone, whereas osteoclasts deficient in either single Rac isoform display normal sealing zone formation [161]. In addition, osteoclasts generated from mice lacking any of three Rac GEFs (Vav3, Dock5 and FARP2) could not assemble sealing zones on bone [162-164]. Interestingly, when activated or dominant-negative Rac1 are microinjected into avian osteoclast-like cells, podosomes disappear [157]. Elongation and branching of F-actin are regulated by WASP, N-WASP, and the Arp2/3 complex, which are downstream of Cdc42 [146]. Studies show that Cdc42 is dispensable for ultimate formation of sealing zones on bone but regulates the rate at which they are generated [165]. In addition, expression of activated Cdc42 leads to podosome dissolution [156,166] in osteoclasts. All these findings reveal the requirement of a tight regulation of Rho GTPases

14 activity during podosomes formation and patterning. Additionally, Rho and Rac exert antagonistic functions on spreading of avian osteoclast-like cells [157]. Cells expressing constitutively activated Rac1 show extensive spreading, whereas cells expressing dominant-negative Rac1 display retraction [157]. The role of Rac in cell spreading is further confirmed by the observations that the Rac GEF Vav3-deficient osteoclasts are incapable of spreading [162]. However, microinjection of constitutively active RhoA triggers rounding up of cells, whereas Rho inhibition by C3 triggers their spreading [157]. Osteoclasts alternate between rounded up morphology with sealing zone during resorption and spread out morphology without any typical actin structures during migration. These cycles of rounding up and spreading out are responsible for osteoclast migration which resembles the movement of an inchworm [150].

1.5.2 Ruffled border and resorption When a resorption compartment is isolated by the sealing zone, protons and proteolytic enzymes are secreted into this compartment to dissolve minerals and degrade bone matrix proteins, respectively (Figure V). The protons are provided by the enzyme carbonic anhydrase + − + II, which generates H and HCO3 from carbonic acid [167]. H is secreted via V-ATPases which consist of V1 and V0 domains. V1 carries out ATP hydrolysis, driving H+ translocation via the V0 domain [168]. Cl− is also secreted via chloride channel ClC-7 to maintain − − electroneutrality [169]. In addition, HCO3 is exported from the cell and Cl is imported into − − the cell by a passive Cl /HCO3 exchanger in the basolateral membrane [170]. The secreted HCl produces a pH of 4.5 in the resorption lacunae, leading to the dissolution of mineral component of bone [168]. The degradation of the organic component of bone matrix is mainly accomplished by a cysteine protease, CTSK. Since the absence of CTSK in osteoclasts resulted in mild osteopetrosis, rather than a fully blocked osteoclastic function, suggesting that other enzymes also contribute to the degradation of organic matrix [171]. MMP-9, a matrix metalloproteinase, is highly expressed in osteoclasts and thought to play an important role in bone resorption [172]. TRAP, a phenotypic marker of osteoclasts, begins to be expressed in the osteoclast precursors before the fusion stage. However, its function has not been elucidated yet [76]. The ruffled border, the resorptive organelle of the osteoclast, is a highly convoluted membrane enclosed by the sealing zone. It forms as a result of extensive fusion of vesicles containing the resorption machinery such as CTSK, V-ATPase and ClC-7 to the bone-apposed plasma membrane (Figure V) [173]. This vesicular trafficking is regulated by two independent

15 pathways, a Rab7-dependent lysosomal pathway and a Rab3D-dependent non-lysosomal pathway [174]. Synaptotagmin VII is a lysosome-associated protein and mediates the fusion of lysosomes with the ruffled border [175]. Bone degradation products are endocytosed at the ruffled border, then transported along a transcytotic vesicular pathway toward the functional secretory domain (FSD) at the center of the bone opposing surface, and finally released [176,177].

1.6 Cell culture models of osteoclastogenesis The Copernican revolution in osteoclast biology began with the identification of M-CSF and RANKL as the key cytokines to induce osteoclast formation [178]. Before the recombinant human and mouse M-CSF and RANKL are available, the only way to study osteoclasts in vitro was to directly isolate mature osteoclasts from neonatal animals such as chicks [179] and rabbits [180] by fragmentation of their bones. Subsequently, a co-culturing system containing osteoclast precursors and osteoblasts under osteoclastogenic conditions was developed to generate osteoclasts [181]. These historical methods are now rarely used, but they may still be the most appropriate methods for certain investigations, such as to determine the effects of pharmacological agents on osteoclasts in vivo, to assess osteoclast mediated bone resorption without affecting osteoclast formation, and to study osteoblast-osteoclast interactions. Nowadays it is possible to generate osteoclasts from primary osteoclast precursors as well as from cell lines such as RAW 264.7 by treating the cells with M-CSF and RANKL. The sources of osteoclast precursors can be bone marrow from long bones of adult mice [182] and human peripheral blood mononuclear cells (PBMCs) from fresh peripheral blood or from buffy coat [183]. Murine bone marrow cells or PBMCs are treated with M-CSF to generate M-CSF- dependent macrophages. Then treatment with M-CSF and RANKL drives the osteoclastogenesis [178]. Alternatively, the RAW 264.7 murine macrophage cell line can be used as osteoclast precursors [184]. The treatment with RANKL alone is sufficient to induce osteoclastogenesis since RANKL augments M-CSF production in RAW 264.7 cells [185]. The passage number of the RAW 264.7 cells can affect osteoclastogenesis, and they can differentiate into osteoclasts only before the passage 20 [186]. Using RAW 264.7 has many advantages over primary cells, such as wide availability, homogeneous nature devoid of osteoblasts and lymphocytes, and ease of transfection [187]. However, immortalized cell lines may not respond to stimuli in the same way as primary cells and should be used in parallel to primary cells [178].

FIGURE IV activity

Podosome cluster Actin patch RhoA

Podosome ring

Sealing zone Podosome belt

Non-bone substrate Bone substrate

RhoA activity

F-actin cloud Actin patch F-actin core

FIGURE IV Scheme of podosome belt and sealing zone formation in cultured osteoclasts. On non-mineralized substrate, podosomes initially pattern into clusters. As the cell matures, podosomes are arranged into short-lived internal rings which expand to the cell periphery to form a stable podosome belt where the podosomes become denser and tightly interconnected. On the other hand, on mineralized substrate, podosomes initially pattern into the actin patches which are uncharacterized actin clusters, and finally develop into a thickened, stable annular structure referred to as sealing zone where the podosomes become more densely packed and more interconnected. Lower Rho activity in cells is required for podosome belt formation. Modified from [177]. 17 FIGURE V

Transcytosis Cl- - HCO3

- Cl - HCO3 CTSK H2O H2CO3 CO2 CO2 Cl-

H+

Sealing zone Cl- Ruffled border H+

Bone: Hydroxyapatite Type I collagen

CTSK

Chloride-bicarbonate Integrin α β Lysosome v 3 exchanger

ClC-7 V-ATPase Degraded products

FIGURE V Bone resorption and ruffled border formation. The integrin αvβ3 are responsible for the adherence of osteoclasts to bone surface. Sealing zone isolates the resorption pit into which osteoclasts secrete H+ and Cl− via proton pump V-ATPase and chloride channel ClC-7. H+ is produced by the dehydrogenation of a carbonic acid. Cl− is imported into osteoclasts in exchange − for HCO3 . Acidification of the resorption pit leads to the decalcification of the bone. Cathepsin K (CTSK) degrades the organic components, mainly type I collagen. The ruffled border is a highly convoluted membrane enclosed by the sealing zone. It forms as a result of extensive fusion of vesicles containing the resorption machinery such as CTSK, V-ATPase and ClC-7 to the bone- apposed plasma membrane. Bone degradation products are endocytosed at the ruffled border, then transported along a transcytotic vesicular pathway toward the functional secretory domain at the center of the bone opposing surface, and finally released. Modified from [46]. 18 1.7 Transglutaminases (TGs) The transglutaminase (protein-glutamine γ-glutamyltransferase, TGase, TG, EC 2.3.2.13) family consists of nine structurally and functionally related proteins, namely, eight enzymes designated TG1-7 and Factor XIII-A (FXIII-A) as well as band 4.2 which lost its catalytic activity in evolution and mainly functions as a structural protein on erythrocyte membranes (Figure VI) [188,189]. TGs catalyze a variety of Ca2+- and thiol-dependent reactions, leading to the post-translational modification of proteins. In addition, some TGs are found to display activities as atypical GTPase and ATPase, protein disulfide isomerase, protein kinase, and as well as other non-enzymatic functions in cell signaling and cell-matrix interactions [190], putting them in an essential position to regulate various biological processes such as extracellular matrix assembly, skin barrier formation, blood coagulation and energy metabolism.

1.7.1 Structures and conformations of TGs

All TGs contain four sequential and structurally distinct domains, i.e., an NH2-terminal β- sandwich (TG1 and FXIII-A have an additional NH2-terminal pro-peptide whose removal promotes their activation), an α/β catalytic core, and two COOH-terminal β-barrel domains. The active site is located in the catalytic core and the catalytic triad contains a cysteine, histidine, and aspartate residue. In addition, a conserved tryptophan is also essential for catalysis via stabilizing the transition state during the formation of acylenzyme intermediate (Figure VII). Band 4.2, however, contains none of these four residues [188]. When bound to GTP, GDP or ATP, TG2 adopts a compact or closed conformation with the active sites occluded by the C-terminal β-barrels [191]. However, upon binding Ca2+, TG2 undergoes a dramatic conformational change in which the C-terminal β-barrels swing down from their start position, conveying TG2 with a linear conformation, making the catalytic core accessible to the substrates [192]. Since the GTP-binding site is located in a cleft between the catalytic core and the first β-barrel, this extended or open conformation abolishes the purine nucleotide binding activity of TG2 (Figure VII) [192]. In addition, some TG inhibitors, such as monodansylcadaverine (MDC) [193], NC9 [194,195], Z006 [192] (see below 1.10), can lock TG2 in the open conformation. It is generally assumed that intracellular TG2 is mainly in the closed conformation due to the relatively high level of GTP/GDP/ATP and relatively low concentration of free Ca2+. Conversely, extracellular TG2 may adopt the open conformation, because of the low concentration of GTP/GDP/ATP and the high level of Ca2+ [191].

FIGURE VI

band 4.2 (inactive)

TG 5 / TG x

TG 7 / TG Z

TG 6 / TG Y

TG 3 / Epidermal TG / TG E

TG 2 / Tissue TG / TG C

TG 4 / Prostate TG / TG P Factor XIIIA-subunit

TG 1 / Keratinocyte TG / TG K Invertebrate TG

FIGURE VI Phylogenetic tree of transglutaminase family. Phylogenetic tree showing protein sequence similarity between members of the human transglutaminase family. Modified from [189].

20 FIGURE VII

a Trp Cys His Asp

β-sandwich Core Barrel 1 Barrel 2

b 2 barrels

1 barrels 1 Ca2+ Ca2+ GTP TG active site core 2 GTP core

β-sandwich β-sandwich

FIGURE VII Protein domains and conformational changes of transglutaminase. (a) The four structural domains of the protein (β-sandwich, core, and barrels 1 and 2) and the relative location of the transition-state-stabilizing Trp, and active site Cys, His, and Asp are shown. (b) Crystal structures of TG2 reveal two different conformational states. When bound to GTP, GDP or ATP, TG2 adopts a compact or closed conformation with the active sites occluded by the C- terminal β-barrels. However, upon binding Ca2+, TG2 undergoes a dramatic conformational change in which the C-terminal β-barrels swing down from their start position, conveying TG2 with a linear conformation, making the catalytic core accessible to the substrates. Since the GTP-binding site is located in a cleft between the catalytic core and the first β-barrel, this extended or open conformation abolishes the purine nucleotide binding activity of TG2. Modified from [188].

21 1.7.2 Enzymatic functions of TGs TGs primarily catalyze six enzymatic reactions which can be divided into three types: transamidation (cross-linking, amine incorporation, and acylation), esterification, and hydrolysis (deamidation and isopeptide cleavage) (Figure VIII) [188]. Among the six reactions, five (except for isopeptide cleavage) share a common step, i.e., the formation of the acylenzyme intermediate (γ-glutamylthioester) between the γ-carboxamide group of a protein- or peptide-bound glutamine residue (also known as the “Q” substrate) and the active site cysteine of the enzyme, and the generation of ammonia as a by-product [188]. Then this thiolester bond is attacked and cleaved by another molecule which can be an ε-amine group of a protein-bound lysine residue or a primary amine (also known as the “K” substrate) (transamidation reaction), or an alcohol (esterification), or water (hydrolysis), leading to the generation of end products and the re-establishment of active site cysteine residue of the enzyme [188,196]. Specifically, when a protein-bound glutamine residue reacts with a protein- bound lysine residue and form an Nε-(γ-glutamyl)lysine isopeptide bond, it leads to cross- linking, conferring rigidity and protease resistance to proteins [188]. When a protein-bound glutamine residue reacts with a primary amine, it results in amine incorporation [188]. When a peptide-bound glutamine residue reacts with a protein-bound lysine residue, it causes the acylation of the protein [188]. When a protein-bound glutamine residue reacts with an alcohol, it leads to esterification of the protein [188]. When a protein-bound glutamine residue reacts with H2O and the -NH2 group of the glutamine residue is replaced by an -OH group, it results in the deamidation [188]. When the isopeptide bond reacts with H2O, it causes isopeptide cleavage and the scission of cross-linked proteins [188].

FIGURE VIIIa

Transamidation

a) Cross-linking O ǁ ̶ CH2CH2C ̶ NH2 + H2NCH2CH2CH2CH2 ̶

O ǁ ̶ CH2CH2C ̶ NCH2CH2CH2CH2 ̶ + NH3 H

b) Amine incorporation O ǁ ̶ CH2CH2C ̶ NH2 + H2NR

O ǁ ̶ CH CH C ̶ NR + NH 2 2 H 3

c) Acylation O ǁ R’CH2CH2C ̶ NH2 + H2NCH2CH2CH2CH2 ̶

O ǁ R’CH2CH2C ̶ NCH2CH2CH2CH2 ̶ + NH3 H

FIGURE VIIIa Transglutaminases catalyze various post-translational reactions. Transamidation can cause a) protein cross-linking by forming a Nε(γ-glutamyl)lysine isopeptide bond between the lysine residue of one protein (purple ellipse) and the glutamine residue of another protein (blue rectangle); b) the incorporation of an amine (H2NR) into the glutamine residue of the protein; and c) the acylation of a lysine side chain of the protein. R represents the side chain in a primary amine. R′ represents a glutamine-containing peptide. Modified from [196].

23 FIGURE VIIIb

Esterification

d) O ǁ ̶ CH2CH2C ̶ NH2 + HOR’’

O ǁ ̶ CH2CH2C ̶ OR’’ + NH3

Hydrolysis

e) Deamidation O ǁ ̶ CH2CH2C ̶ NH2 + H2O

O ǁ ̶ CH2CH2C ̶ OH + NH3

f) Isopeptide cleavage O ǁ R’’’CH CH C ̶ NCH CH CH CH R’’’’ + H O 2 2 H 2 2 2 2 2

O ǁ R’’’CH2CH2C ̶ OH + H2NCH2CH2CH2CH2R’’’’

FIGURE VIIIb Transglutaminases catalyze various post-translational reactions. d) When a protein-bound glutamine residue reacts with an alcohol, it leads to esterification of the protein. e) When a protein-bound glutamine residue reacts with H2O and the -NH2 group of the glutamine residue is replaced by an -OH group, it results in the deamidation. f) When the isopeptide bond reacts with H2O, it causes isopeptide cleavage and the scission of cross-linked proteins. R′′ represents a ceramide. R′′′ and R′′′′ represent the side chains in branched isopeptides. Modified from [196]. 24 1.8 Regulation of TG activity TG activity can be regulated not only by Ca2+, GTP/GDP/ATP and proteolysis, but also by redox status (oxidative stress and nitrosative stress) [197], although, this has been mostly established for TG2. Reactive oxygen species (ROS) are generated during mitochondrial oxidative metabolism as well as in response to various stimuli in the surrounding environment [198]. On the other hand, there are a variety of antioxidant systems present in cells, such as thioredoxin and glutathione, to eliminate ROS. When the balance between ROS and antioxidants tips towards excessive ROS, oxidative stress occurs [199]. Reduced glutathione/oxidized glutathione (GSH/GSSG) ratio is a major determinant of oxidative stress [200]. Human TG2 has 20 cysteines in total, including one cysteine (Cys277) in the active site, making it susceptible to oxidation [201]. Among all the post-translational modifications mediated by ROS, disulfide bond formation is the most studied reaction. Sollid group identified a cysteine triad consisting of Cys230, Cys370, and Cys371 involved in oxidative inactivation of TG2. Cys370 can participate in disulfide bond formation with both Cys230 and Cys371, where the formation of the Cys230-Cys370 appears to facilitate the formation of the Cys370- Cys371. These disulfide bonds inhibit the catalytic activity of TG2 but maintain TG2 in the open conformation. However, it is currently unclear how the disulfide bonds block the catalytic activity of TG2 in the open conformation [202]. This oxidative inactivation of TG2 is confirmed by the inactive form of TG2 in highly oxidative state in the extracellular space [203]. However, the disulfide-bonded extracellular TG2 can be re-activated by thioredoxin-mediated reduction [204]. Nitric oxide (NO), one of the most studied reactive nitrogen species (RNS), is a relatively stable free radical produced during inflammation and nitrosylates free cysteines and forms nitrosothiol (RSNO), altering the biological activity of proteins [205]. It is well known that TG2 can be modified by NO, leading to the S-nitrosylation of key cysteine residues and then the inhibition of its catalytic activity [206,207]. Furthermore, in aging vasculature, the reduced NO production can lead to decreased S-nitrosylation but increased externalization of TG2, and increased extracellular matrix cross-linking and vascular stiffness [208]. In addition to cysteine residues, NO was reported to increase the nitrosylation of tyrosine residues in TG2 [209]. Furthermore, there is a tyrosine residue (Tyr274) critical for activation in human TG2 [210]. However, the biological roles of tyrosine nitrosylation in TG2 remain to be clarified.

1.9 TG-specific substrate peptides Despite the fact that TGs share the common catalytic reaction to form the acylenzyme intermediate, they generally exhibit remarkable specificity for the glutamine-donor substrates and this specificity is governed by the primary sequence and/or tertiary structure around the reactive glutamine residues. Hitomi group made a major breakthrough to identify the preferred glutamine-containing sequences for TG1, TG2, TG3, TG6, TG7, and FXIII-A by screening through the M13 phage-displayed peptide library (Figure IX) [211-216]. These peptides, containing only 12 amino acid residues, act as the glutamine-donor substrates with high reactivity and isozyme specificity, allowing to evaluate the individual TG activity. When the glutamine residue (Q) is replaced by an asparagine residue (N) to abrogate its reactivity to TGs, it can be used as a negative control. In addition, biotin group can also be added into the peptides for various detection purposes. These peptides are widely used in evaluation of individual TG activity in microtiter plate-based assay [217,218], visualization of in situ TG-specific activity [211,214,215,219,220], as well as the detection, purification and identification of the lysine- containing substrates in a TG-specific manner [221-223]. However, TGs have a less specificity for the lysine-donor substrate. Hitomi group failed to obtain specific lysine-donor peptides by using the same screening method. Therefore, the principle lysine-donor substrate in use, 5- (biotinamido)pentylamine (bPA), is a primary amine coupled with biotin to evaluate the total TG activity.

1.10 TG inhibitors The broad spectrum of pathologies in which TGs are implicated suggests that they may be interesting therapeutic targets, however, this necessitates the development of potent, yet very specific TG inhibitors/modulators. The development of such compounds has been an active area and resulted in large number of inhibitors with different modes of action. The TG inhibitors can be divided into three categories: pseudo-substrates, peptidomimetics, and novel scaffolds [224]. The inhibitors used in this thesis are MDC, NC9, Z006 and T101 (Figure X). The pseudo-substrate inhibitors were identified in the early 90’s. Among these inhibitors, MDC is generally considered to be the most potent [225]. It contains a primary amine and act as an amine-donor substrate for TGs, preventing the cross-linking from occurring. In addition to inhibit TGs, MDC also displays inhibitory activity against 17-ketosteroid reductase and phosphatidylcholine synthetase [225]. MDC has intrinsic fluorescence due to the fluorescent dansyl group, giving it labeling capability. MDC is a permeable inhibitor and is widely used in

26 cell culture systems [224]. However, it is not successful in the clinic due to its poor tolerance in mice [225]. The peptidomimetic inhibitors are the best developed category so far. Most of them are irreversible inhibitors, containing a peptidomimetic scaffold and an electrophilic warhead adjacent to the nucleophilic thiol (cysteine) in the active site of TGs, generating a covalent bond [226]. The compound NC9, that was used in the work described in this thesis, was first described in 2008 and carries an acrylamide warhead that reacts with active cysteine in TGs and remains irreversibly, covalently bound to the enzyme [227]. NC9 is moderately cell permeable [194]. Z006 utilizes a reactive 6-diazo-5-oxo-L-norleucine functionality to bind active cysteine in TGs [224]. It becomes cell permeable at 40 µM [228]. The third category of TG inhibitors contains novel scaffolds which do not function as alternative substrates for TGs. T101 works via the 2-[(2-oxopropyl)thio]imidazolium functionality. It is cell impermeable [224].

FIGURE IX

Transglutaminase-specific substrate Sequence peptides bK5 (TG1) Biotinyl-YEQHKLPSSWPF

bT26 (TG2) Biotinyl-HQSYVDPWMLDH

bE51 (TG3) Biotinyl-PPPYSFYQSRWV

bY25 (TG6) Biotinyl-DDWDAMDEQIWF

bZ3S (TG7) Biotinyl-YSLQLPVWNDWA

bF11 (FXIII-A) Biotinyl-DQMMLPWPAVAL

bPA (All TGs) 5-(biotinamido)pentylamine

FIGURE IX Sequences of transglutaminase-specific substrate peptides. Peptides K5 (TG1), T26 (TG2), E51 (TG3), Y25 (TG6), Z3S (TG7), and F11 (FXIII-A) contain only 12 amino acid residues, act as the glutamine-donor substrates with high reactivity and isozyme specificity, allowing to evaluate the individual TG activity. In addition, biotin group can also be added into the peptides for various detection purposes. TGs have a less specificity for the lysine-donor substrate. Therefore, the principle lysine-donor substrate in use, 5- (biotinamido)pentylamine (bPA), is a primary amine coupled with biotin to evaluate the total TG activity. Sequences from Zedira.

28 FIGURE X

NC9 MDC

Z006 T101

FIGURE X Structures and function groups of transglutaminase inhibitors. NC9 carries an acrylamide warhead that reacts with active cysteine in TGs and remains irreversibly, covalently bound to the enzyme. MDC contains a primary amine and act as an amine-donor substrate for TGs, preventing the cross-linking from occurring. Z006 utilizes a reactive 6- diazo-5-oxo-L-norleucine functionality to bind active cysteine in TGs. T101 works via the 2- [(2-oxopropyl)thio]imidazolium functionality. Structures from Zedira.

29 1.11 Transglutaminase family members 1.11.1 Transglutaminase 2 (TG2) TG2 is a 76-80 kDa protein and is also named as tissue TG due to its ubiquitous expression in cells and tissues. Mainly a cytosolic protein, it is also found in the nucleus, on the plasma membrane, as well as in the extracellular matrix [229]. Studies show that P2X7 receptor, NO, and perinuclear recycling endosomes regulate TG2 externalization [230-233] and that TG2 binding to cell surface heparan sulfate proteoglycans is critical for its translocation to the cell surface [234]. TG2 is an extremely versatile protein, exhibiting cross-linking/deamidation, GTPase, ATPase, protein disulfide isomerase (PDI), protein kinase, and scaffold function [190]. The cross-linking activity of TG2 plays an essential role in the tissue remodeling/wound healing and extracellular matrix assembly. TG2 is involved in the pathogenesis of various diseases, such as neurodegenerative diseases, liver cirrhosis, renal scarring and pulmonary fibrosis [235]. TG2 also induces deamination of gluten-derived gliadin peptides which are involved in the pathogenesis of celiac disease [236]. Since TG2 exhibits GTPase activity, it can also function as a G protein (Ghα), independent of its transamidation activity, participating in G protein coupled receptors (α1 adrenergic receptors) mediated signaling pathways [237]. Without agonist binding, calreticulin/Ghβ is tightly associated with GDP-bound TG2/Ghα and inhibits both its GTP binding/hydrolytic and transamidation activities. Receptor activation by agonist facilitates the exchange of GDP to GTP and the subsequent dissociation of GTP-bound TG2/Ghα from calreticulin/Ghβ [238]. This activation stimulates PLCδ1, resulting in phosphoinositide hydrolysis and increased intracellular Ca2+. Signaling by TG2/Ghα is terminated by hydrolysis of GTP to GDP and re- association with free calreticulin/Ghβ [238]. In addition to GTP, TG2 was found to bind and hydrolyze ATP. In contrast to GTP hydrolysis, the ATPase activity of TG2 is resistant to Ca2+ in extracellular space [239]. Furthermore, the ATPase activity can be elevated via proteolysis by membrane-type matrix metalloproteinase-1 (MT1-MMP) [240]. The PDI activity of TG2 requires free sulfhydryl groups and is independent of Ca2+ and GTP. An important role for the PDI activity of TG2 is to regulate the correct assembly and function of the mitochondrial ADP/ATP transporter adenine nucleotide translocator 1 (ANT1) [241]. TG2 can also exert an intrinsic serine/threonine protein kinase activity and phosphorylate a variety of proteins such as insulin-like growth factor-binding protein-3 (IGFBP3) and p53 tumor suppressor protein [242,243]. Additionally, TG2 has the scaffolding function on cell surface through a noncovalent

30 association with the extracellular domains of β1 and β3 integrin and the formation of stable ternary complexes with both integrin and fibronectin (FN), promoting the cell adhesion [244]. In addition, TG2 can also regulate cell adhesion and outside-in signaling by via its non-covalent interactions with syndecan-4, growth factor receptors, and other cell surface or extracellular matrix proteins [245]. In spite of the multifunction of TG2, TG2 null mice (Tgm2-/-) do not exhibit any evident abnormality under normal physiological conditions [246]. However, when cell death is induced, Tgm2-/- mice display defective clearance of apoptotic cells by macrophages both in the thymus and liver. Tgm2-/- mice also show inflammatory infiltrates at the apoptosis sites in short term and then developed autoimmunity in the long term [247,248]. In addition, the TG2 deficient mice show impaired insulin secretion and thus glucose intolerance and hyperglycaemia [249]. TG2 deficient cardiomyocytes are also more sensitive to ischemia/reperfusion injury [250]. TG2 deficient mice developed significantly reduced pulmonary fibrosis compared with wild- type mice after bleomycin challenge [251].

1.11.2 Factor XIII-A (FXIII-A)

FXIII circulates in the plasma as heterotetramers (FXIII-A2B2), consisting of two catalytic subunits (FXIII-A2) and two inhibitory subunits (FXIII-B2) [252]. Mature FXIII-A contains 731 amino acids and its molecular weight is 83 kDa [253]. The first 37 amino acids of the N- terminus form the so called activation peptide (AP-FXIII) which buries the active site cysteine to maintain the FXIII in inactive state. Therefore, its cleavage between Arg37 of AP-FXIII and Gly38 of the β-sandwich domain by thrombin is necessary for FXIII-A activation in coagulation process [254]. After removal of the AP-FXIII, its molecular weight decreases by 4 kDa [255]. In the catalytic core domain, Cys314 is the active site of the enzyme [254]. In addition to AP-FXIII, Cys314 is also buried by Tyr560 side chain in β-barrel 1. Thus, the dislocation of Tyr560 is also crucial for activation of the FXIII-A [254]. Furthermore, Ala457 of the core domain is the main binding site for Ca2+ [254]. In addition, Arg260 in the catalytic core domain forms a salt bridge with Asp404 from another catalytic core domain to form the FXIII-A homodimer [254]. Mature FXIII-B consists of 641 amino acids and it is synthetized by hepatocytes [256]. The activation of FXIII is initiated by thrombin, leading to the cleavage of AP-FXIII. In the presence of Ca2+, the inhibitory B subunits dissociate, and the cleaved FXIII-A dimer adopts an enzymatically active conformation after an opening of the structure, leading to the fully activated enzyme [254]. As the last zymogen activated in the blood

31 coagulation cascade, FXIII-A catalyzes the cross-linking of fibrin and stabilizes the fibrin clot. Thus, FXIII-A is crucial to maintain hemostasis and its deficiency causes severe bleeding diathesis [253]. Platelets were proposed to be the source of FXIII-A in the plasma, however, recent study shows tissue resident macrophages are in fact the major producer of plasma FXIII- A [257]. The cellular form of FXIII is a homodimer of subunit A, and it is present in platelets, monocytes, macrophages, dendritic cells, chondrocytes, osteoblasts, osteocytes, and adipocytes [256]. Low Ca2+ concentration is sufficient to induce a slow non-proteolytic activation of cellular FXIII-A at physiological ionic strength and this activation rate is greatly elevated by increasing the ionic strength [258]. In contrast to cellular FXIII-A, an extremely high Ca2+ concentrations (≥100 mM) is required for the non-proteolytic activation of plasma FXIII where the A2B2 complex dissociates and the non-truncated A2 becomes active [253]. In the primary structure of FXIII- A, there is no identifiable hydrophobic leader sequence, therefore, it is secreted via the nonclassical ER-Golgi pathway [256]. Studies show that FXIII-A in macrophages associates with podosomes and other structures adjacent to the plasma membrane. Furthermore, FXIII-A is found in intracellular vesicles positive for Golgi matrix protein 130, which has been implicated in the trafficking of non-classically secreted proteins to the plasma membrane [259]. In addition to its role in hemostasis, multifunctional FXIII-A also has a pivotal role in maintenance of pregnancy, angiogenesis, wound healing, and bone and energy metabolism [254,260]. FXIII- A deficient mice developed severe uterine bleeding events and subsequent spontaneous miscarriages [261], and suffered from impaired wound healing and fatal rupture of the left ventricle after myocardial infarction [262,263]. In the Matrigel plug model, FXIII-A deficient mice showed a significantly decreased number of new vessels compared to wild type [264]. FXIII-A knockout mice also developed severe signs of inflammation at the site of infection in a streptococcal skin infection model [265]. Recent studies demonstrated that FXIII-A deficient mice showed characteristics of metabolically healthy obesity with improved insulin sensitivity on high fat diet [260].

1.11.3 Transglutaminase 1 (TG1) TG1, also known as keratinocyte TG, is mainly expressed in the stratified squamous epithelia of the integument, the upper digestive tract, and the lower female genital tract [190]. TG1 is synthesized as a 92 kDa protein, and turned into a 106 kDa form after post-translational modifications [266]. These modifications include one constitutive N-myristoylation at Gly3

32 and two S-myristoylation at Cys residues in proliferating epidermal keratinocytes or one S- palmitoylation at Cys residue in differentiating epidermal keratinocytes, all of which lie in the N-terminal region [267]. TG1 is unique among TGs since it can be anchored to the plasma membrane via the above-mentioned S-acylated 10 kDa N-terminal region [266]. The membrane-bound TG1 exists in two forms: a full-length inactive zymogen, or a complex consisting of 67/33/10 kDa fragments which is highly active [268]. The cytosolic TG1 exists as a full-length form with low activity, or two proteolytically processed forms, i.e., one 67 kDa fragment and a complex consisting of 67/33 kDa fragments, both of which have higher activities than the full-length form [269]. This 67 kDa fragment contains the catalytic domain and is generated by two proteolytic cleavages between Arg93 and Gly94 and between Arg573 and Gly574 [270]. Studies show that cathepsin D plays an essential role in processing the 150 kDa TG1 precursor to an enzymatic active 35 kDa form in vivo [271]. Studies have shown that guanine nucleotides do not inhibit TG1 activity [272]. TG1 is required for the formation of cornified cell envelope that acts as protective skin barrier. Mutations in human TGM1 cause lamellar ichthyosis in humans, an autosomal recessive skin disorder [190]. TG1 knockout mice were developed and the Tgm1-/- neonates were smaller with a lower body weight and had erythrodermic skin with abnormal keratinization. They also displayed defective cell envelope assembly in the stratum corneum and thus impaired skin barrier formation. In addition, these mice died within 4-5 hr after birth due to progressive dehydration [273].

1.11.4 Other Transglutaminases Transglutaminase 3, also known as epidermal TG, is found in hair follicles, epidermis, and brain. Similar to TG2, TG3 can also bind to and hydrolyze GTP [190]. It is required for hair fiber morphogenesis [274]. Transglutaminase 4, also referred to as prostate TG, is a prostate-specific transglutaminase and present in the seminal plasma [190]. It is required to form the copulatory plug in mice and the defects in copulatory plug formation lead to defective fertility [275]. TG4 is also a prospective marker of prostate cancer progression [276]. Transglutaminase 5 is widely expressed in the epidermis and the loss-of-function mutations of TGM5 result in skin peeling syndrome in humans [277]. It is also involved in the hyperkeratosis in ichthyosis and psoriasis patients [278]. The cross-linking activity of TG5 can be inhibited by GTP and ATP [190]. Transglutaminase 6 and transglutaminase 7 have a similar expression pattern which is restricted

33 to testes, lungs, and brain [190]. Mutations in TGM6 cause spinocerebellar ataxia type 35, a rare autosomal dominant neurodegenerative disease [279]. The function of TG7 remains unknown.

1.12 TGs in monocytes and macrophages For detailed review on expression and regulation of TGs in macrophages and on the distinct functions of TGs in monocytes and tissue-specific macrophages, please refer to Chapter 2.

1.13 TGs in bone remodeling It has been long reported that TG2 and FXIII-A are expressed in osteoblasts [280,281]. TG activity plays a critical role in the formation of the FN-COL I network during the early stage of osteoblast differentiation. Disrupting this with TG inhibition affects osteoblast differentiation, matrix accumulation, and the subsequent mineralization [281]. TG2-mediated cross-linking of FN [282], OPN, and BSP [283] significantly promotes osteoblast adhesion. In addition, a truncated 56 kDa fragment of TG2, generated by MT1-MMP proteolysis, acts as an ATPase in a Ca2+-rich environment, promoting matrix mineralization in osteoblasts [239,240]. However, accumulating evidence indicates that FXIII-A, rather than TG2, functions as the major regulator of osteoblast matrix deposition. FXIII-A is found as cellular and extracellular matrix forms in osteoblasts and extracellular COL I can regulate FXIII-A protein levels, cellular localization and secretion [284]. Secreted osteoblast FXIII-A is required for the incorporation of plasma FN into the bone matrix, providing a scaffold for normal COL I deposition [285]. In addition, surface-associated osteoblast FXIII-A is required for the interaction between microtubules and the plasma membrane, promoting COL I secretion [286]. Despite their importance for osteoblast function in vitro, neither Tgm2-/- mice nor F13a1-/- mice show an overt bone phenotype [246,287]. Therefore, to examine if they have synergistic functions in the bone, we generated a mouse lacking both TG2 and FXIII-A (Tgm2-/-;F13a1- /-) [288]. The double knockout, Tgm2-/-;F13a1-/- mice, show compromised biomechanical properties in long bones and dramatic trabecular bone loss in axial and appendicular skeleton which was caused by increased bone resorption and increased osteoclastogenesis in vivo. The phenotype was cell-autonomous and Tgm2-/-;F13a1-/- monocytes retained their increased potential to form osteoclasts when isolated from mice and cultured in vitro. Interestingly, this increased osteoclastogenesis was reversed by chemical inhibition of TG activity with NC9, revealing that there was a remnant TG that was active in osteoclasts and that it was responsible

34 for the increased osteoclastogenesis potential. This TG was TG1 and this was first demonstration of its involvement in this cell type and suggested a potentially important role in bone. TG1 expression is not upregulated in the absence of TG2 and FXIII-A, suggesting that it does not compensate for the absence of TG2 and FXIII-A but has a bona fide role in maintenance of bone mass. Based on these results, it appears that in contrast to TG2 and FXIII- A which synergistically inhibit osteoclastogenesis, TG1 functions to promote it [288]. In addition, there are another two studies on their roles in osteoclastogenesis. The work of Raghu et al. shows that FXIII-A deficiency results in reduced osteoclastogenesis in vivo and in vitro [289]. The work of Kim et al. shows that TG2 deficiency gives rise to increased osteoclastogenesis in vivo and in vitro [290]. As for bone formation axis, TG2/FXIII-A null mice have normal osteoblast activity, but increased bone marrow adipogenesis and increased RANKL production. BMMSCs lacking both TG2 and FXIII-A show higher adipogenic potential, increased RANKL expression and upregulated TG1 expression. In addition, chemical inhibition of TG activity in TG2/FXIII-A null BMMSCs further enhances their adipogenic potential and RANKL expression. The enhanced adipogenesis of TG2/FXIII-A null BMMSCs is associated with plasma FN assembly defect in vitro. TG2/FXIII-A null mice also show FN retention in serum and marrow in vivo due to its defective integration into bone [288]. Collectively, these findings demonstrate that TG2, FXIII-A, and TG1 are novel regulators of bone mass, plasma FN homeostasis, RANKL production, as well as differentiation of monocyte/macrophage lineage cells and BMMSCs [288].

1.14 Hypothesis, rationale and objectives The following two hypotheses are investigated in this thesis: Hypothesis 1 - TG activity regulates osteoclastogenesis at different stages. Rationale - Our work has shown that three TGs are present in osteoclasts and their precursor cells and TG2/FXIII-A deficiency displays increased potential to form osteoclasts in vivo and in vitro. However, chemical inhibition of TG activity with NC9 completely blocks osteoclast formation in both wild type and TG2/FXIII-A deficient monocytes, suggesting TG activity is required for osteoclastogenesis [288]. Osteoclastogenesis is a complex stepwise process that includes defined stages, such as osteoclast precursor commitment, differentiation, multinucleation, resorption, and a dynamic cytoskeletal organization concomitant with the migration, multinucleation and resorption [16]. Here we hypothesize that TG activity regulates

35 osteoclastogenesis at different stages. Objective - We aim to examine effects of TG inhibition on osteoclastogenesis at different stages and to explore the mechanism by which TG activity regulates osteoclast formation.

Hypothesis 2 – Expression and activities of TGs are regulated by M-CSF and RANKL during osteoclastogenesis. Rationale - Our study for TG2/FXIII-A deficient mice has shown that TG2 and FXIII-A inhibit osteoclastogenesis, whereas TG1 promote this process [288]. In addition, there are another two studies on their roles in osteoclastogenesis. The work of Raghu et al. shows that FXIII-A deficiency results in reduced osteoclastogenesis in vivo and in vitro [289]. The work of Kim et al. shows that TG2 deficiency gives rise to increased osteoclastogenesis in vivo and in vitro [290]. In addition, our recent study shows that the expression of TG1, TG2, and FXIII-A display distinct patterns during M-CSF-induced macrophage differentiation and M- CSF/RANKL-induced osteoclast differentiation [291]. All these findings suggest that three TGs may exert distinct functions at different stages of osteoclast formation via different mechanisms. Here we hypothesize that the expression and activities of TGs are differentially regulated by M-CSF and RANKL during osteoclastogenesis. Objective - We aim to examine the mRNA and protein levels of TG1, TG2, and FXIII-A and detect their activities with TG-specific “Hitomi peptides” during osteoclastogenesis.

To test these hypotheses, I have used bone marrow-derived macrophages as osteoclast precursors in vitro in both Chapters 3 and 4 of this thesis. Experiments in Chapters 3 and 4 are designed to test Hypothesis 1 and 2. Chapter 2 is a detailed review on TGs in monocytes and macrophages, which act as osteoclast precursors. Each chapter is a paper manuscript either published or submitted to a peer-review journal.

Chapter 2 - Transglutaminases in monocytes and macrophages

Osteoclasts are derived from monocyte/macrophage lineage cells and numerous studies have shown that TG2 and FXIII-A are extensively involved in monocyte- and macrophage-mediated physiological and pathological processes. Therefore, we outline the current knowledge of TGs in monocytes and macrophages to gain more insight into their role in osteoclasts. Firstly, we review the role of TGs in the adhesion and extravasation of monocytes, the expression and regulation of TGs in macrophages, as well as the role of TGs in the generic function of macrophage, that is, phagocytosis. In addition, we also outline the existing data on the distinct roles of TGs in tissue-specific macrophages, including monocytes/macrophages in vasculature, alveolar and interstitial macrophages in lung, microglia and infiltrated monocytes/macrophages in central nervous system, and osteoclasts in bone. This review greatly deepens our understanding of the roles of TGs in monocytes and macrophages which act as osteoclast precursors, providing more novel perspectives for us to study TGs in osteoclasts.

The study presented in this chapter has been published in the peer-reviewed journal Medical Sciences. Vol 6, Issue 4, pii: E115, 2018  MDPI

Transglutaminases in monocytes and macrophages

Sun H1 and Kaartinen MT1,2 1 Division of Biomedical Sciences, Faculty of Dentistry, McGill University, Montreal, QC H3A 0C7, Canada 2 Division of Experimental Medicine, Department of Medicine, Faculty of Medicine, McGill University, Montreal, QC H3A 0C7, Canada

2.1 Abstract Macrophages are key players in various inflammatory disorders and pathological conditions via phagocytosis and orchestrating immune responses. They are highly heterogeneous in terms of their phenotypes and functions by adaptation to different organs and tissue environments. Upon damage or infection, monocytes are rapidly recruited to tissues and differentiate into macrophages. Transglutaminases (TGs) are a family of structurally and functionally related enzymes with Ca2+-dependent transamidation and deamidation activity. Numerous studies have shown that TGs, particularly TG2 and Factor XIII-A, are extensively involved in monocyte- and macrophage-mediated physiological and pathological processes. In the present review, we outline the current knowledge of the role of TGs in the adhesion and extravasation of monocytes, the expression of TGs during macrophage differentiation, and the regulation of TG2 expression by various pro- and anti-inflammatory mediators in macrophages. Furthermore, we summarize the role of TGs in macrophage phagocytosis and the understanding of the mechanisms involved. Finally, we review the roles of TGs in tissue-specific macrophages, including monocytes/macrophages in vasculature, alveolar and interstitial macrophages in lung, microglia and infiltrated monocytes/macrophages in central nervous system, and osteoclasts in bone. Based on the studies in this review, we conclude that monocyte- and macrophage-derived TGs are involved in inflammatory processes in these organs. However, more in vivo studies and clinical studies during different stages of these processes are required to determine the accurate roles of TGs, their substrates, and the mechanisms-of-action.

2.2 Introduction Macrophages, which mean “big eaters” in Greek, are immune cells present in almost all tissues of the body with distinct tissue-specific phenotypes and functions. The highly heterogeneous macrophages have multiple origins: yolk sac, fetal liver, and circulating monocytes [292].

Macrophages play a central role in tissue homeostasis as well as in inflammation via phagocytosis and orchestrating immune responses [27]. Macrophages kill the ingested pathogens via phagocytosis and protect the organism from infection. Macrophages can also remove apoptotic cells by efferocytosis, a specialized form of phagocytosis, to prevent the induction of inflammation by the apoptotic cells [293]. In addition to their role as guardians, macrophages play an essential role in organizing inflammatory reactions by producing pro- and anti-inflammatory mediators and effector molecules such as chemokines, cytokines, and growth factors [27]. Consistent with such a broad array of functions, macrophages are involved in various inflammatory disorders and pathological conditions, such as atherosclerosis, osteoporosis, neurodegenerative disorders, and autoimmune diseases [294]. The transglutaminase (protein glutamine-γ-glutamyltransferase, TGase, TG, EC 2.3.2.13) family consists of nine structurally and functionally related proteins. Of the nine TGs, one (erythrocyte membrane protein band 4.2) lost its catalytic activity in evolution and acts as a structural protein, and eight (TG1-7 and Factor XIII-A (FXIII-A)) are active enzymes showing the Ca2+-dependent transamidation and deamidation activity [196]. In addition to this enzymatic activity, some TGs may function as atypical GTPase and ATPase, protein disulfide isomerase, protein kinase, and also have other non-enzymatic functions in cell signaling and cell-matrix interactions [190,238,245]. Numerous studies have shown that TGs, particularly TG2 and FXIII-A, are extensively involved in inflammatory processes, making them useful markers for disease diagnosis or potential therapeutic targets [235,295,296]. In this review, we outline the existing data on expression and regulation of TGs in macrophages and elaborate on the distinct functions of TGs in monocytes and tissue-specific macrophages to improve our understanding of their contributions to various diseases.

2.3 Brief Overview of Transglutaminase Family Members Transglutaminase 2, also named as tissue TG due to its ubiquitous expression in cells and tissues, is the most studied member of the family. Mainly a cytosolic protein, it is also found in the nucleus, on the plasma membrane, as well as in the extracellular matrix [297]. Studies show that P2X7 receptor, nitric oxide, and perinuclear recycling endosomes regulate TG2 externalization [230-233] and that TG2 binding to cell surface heparan sulfate proteoglycans is critical for its translocation to the cell surface [234]. Being an extremely versatile protein, TG2 exhibits transamidation, GTPase, ATPase, protein kinase, and protein disulfide isomerase activity. It is involved in various disease processes such as celiac disease and neurodegenerative

39 disorders [190]. As the last zymogen activated in the blood coagulation cascade, FXIII-A catalyzes the stabilization of newly formed fibrin network. Thus, FXIII-A is crucial to maintain hemostasis and its deficiency results in severe bleeding diathesis. It is present in platelets, monocytes, macrophages, dendritic cells, chondrocytes, osteoblasts, osteocytes, and adipocytes [253,298]. Recent study shows resident macrophages are the major source of plasma FXIII-A [257]. In addition to its role in hemostasis, FXIII-A also has a pivotal role in maintenance of pregnancy, angiogenesis, wound healing, and bone and energy metabolism [254,260]. Transglutaminase 1, also known as keratinocyte TG, is primarily expressed in the stratified squamous epithelia of the integument, the upper digestive tract, and the lower female genital tract. It is required for the formation of cornified cell envelope that acts as protective skin barrier [190]. Mutations in TGM1 cause lamellar ichthyosis in humans, an autosomal recessive skin disorder [299-301]. Transglutaminase 3, also referred to as epidermal TG, is found in hair follicles, epidermis, and brain. Analogous to TG2, TG3 can also bind to and hydrolyze GTP [190]. It is required for hair fiber morphogenesis [274]. Transglutaminase 4, also known as prostate TG, is a prostate-specific transglutaminase and present in the seminal plasma [190]. It is required to form the copulatory plug in mice and the defects in copulatory plug formation lead to reduced fertility [275]. TG4 is also a prospective marker of prostate cancer progression [276]. Transglutaminase 5 is widely expressed in the epidermis and the loss-of-function mutations of TGM5 result in skin peeling syndrome in humans [277]. It is also involved in the hyperkeratosis in ichthyosis and psoriasis patients [278]. The cross-linking activity of TG5 can be inhibited by GTP and ATP [190]. Transglutaminase 6 and TG7 have a similar expression pattern which is restricted to testes, lungs, and brain [190]. Mutations in TGM6 cause spinocerebellar ataxia type 35, a rare autosomal dominant neurodegenerative disease [279]. The function of TG7 remains unclear. Some studies suggest that TGs may be able to compensate for each other’s functions partially in tissues where their expressions and activities overlap [302-305].

2.4 Development and Classification of Macrophages Blood monocytes were historically considered as a reservoir for macrophages [306,307], however, in recent years, this paradigm has shifted dramatically [20]. At steady state (i.e.,

40 without inflammatory cues), the contribution of monocytes to macrophages is only restricted to a few specific tissues including the intestine, dermis, heart, and pancreas [308]. Numerous studies have determined that some macrophages and their precursors are embryonically established in the yolk sac and fetal liver before the generation of the first hematopoietic stem cells (HSCs) [309-314]. Tissue macrophages can maintain themselves through self-renewal during adulthood with minimal monocyte input [315-319]. Development, proliferation, differentiation, and function of macrophages are regulated by two growth factors: colony stimulating factor 1 (CSF-1), also known as macrophage-colony stimulating factor (M-CSF), and interleukin-34 (IL-34) [29]. Macrophages are distinguished as large vacuolar cells and they are phenotypically defined positive for F4/80 in mice [28]. Macrophages can be classified based on their tissue of residence and activation. Governed by tissue specific cues, macrophages can adapt to different milieus, such as splenic macrophages, peritoneal macrophages, alveolar and interstitial macrophages (lung), Kupffer cells (liver), osteoclasts (bone), and microglia (central nervous system, CNS) [29]. Macrophages can also respond to different activation signals (cytokines or bacterial products) with changes in their morphology, phenotype, transcriptome, and function. Macrophage activation occurs in two modes, i.e., classically activated (M1) and alternatively activated (M2) macrophages. M1 is activated by lipopolysaccharide (LPS)/interferon gamma (IFN-γ) and exhibits pro-inflammatory features, whereas M2 results from IL-4/IL-13 stimulation and displays anti-inflammatory properties [30]. M1 and M2 perform diametrically opposed functions by generating either nitric oxide (NO) or ornithine via different cleavages of terminal nitrogen linkages of arginine. M1 performs killing/inhibitory functions by producing NO, which inhibits cell proliferation and kills microbes, while M2 displays healing/repair capacities via generating ornithine, which promotes cell proliferation and repair [320].

2.5 Transglutaminases in Monocyte Adhesion and Extravasation As mentioned above, the majority of macrophages are seeded before birth and exhibit negligible need for monocytes throughout adulthood due to their ability of self-proliferation [308]. However, monocyte recruitment into tissues and monocyte-derived macrophages are strongly increased during inflammation [21]. In order to reach the inflamed sites and exploit their functions, cell adhesion and extravasation are required for circulating monocytes. Over 30 years ago, TG2 and FXIII-A were found present in monocytes [321,322]. Akimov et al.

41 demonstrated that both TG2 knock-down and inhibition of TG2-fibronectin binding by function-blocking antibodies decreased the adhesion and spreading of monocytes on fibronectin [323]. In addition, antibodies against β1 and β3 integrins resulted in the similar inhibition but in a less prominent manner. Notably, antibody against FXIII-A had no such effects, suggesting that only TG2 is involved in adhesion and spreading of monocytes [323]. In addition, studies showed TG2 expression was highly upregulated in monocytes during their adhesion onto endothelial cells, indicating that TG2 is required for monocyte extravasation [324]. However, in another study FXIII-A was shown to have a role in monocytes, where it catalyzed the dimerization of angiotensin II type 1 receptor (AT1 receptor) on monocytes. This FXIII-A-dependent dimeric form of the AT1 receptor displayed enhanced signaling and increased monocyte adhesion to endothelial cells in hypertensive patients [325]. Therefore, both TG2 and FXIII-A contribute to the monocyte adhesion and extravasation but via different mechanisms.

2.6 Transglutaminases in Macrophages Expression and Regulation of Transglutaminases in Macrophages Over 30 years ago, TG2 and FXIII-A were also found present in macrophages [326-329] and both of them were identified as novel markers for alternatively activated M2 macrophages [330,331]. Numerous studies demonstrated that level of TG2 increased dramatically during the differentiation from monocytes to macrophages, induced either by adherence to the cell culture dish or by several macrophage stimulating factors such as serum retinoids, IFN-γ, LPS, and 12-O-tetradecanoyl-phorbol-13-acetate (TPA) [321,332-334]. In addition, various stimulating factors participate in the regulation of TG2 expression in differentiated macrophages. Numerous studies demonstrated that retinoic acid was able to induce TG2 expression in macrophages [335-338] and this induction involved cyclic AMP and protein kinase C- dependent pathway [339,340]. Ghanta et al. showed that LPS was a potent inducer of TG2 expression in macrophages and this induction was co-regulated by metastatic tumor antigen 1 (MTA1) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling [341]. In response to LPS stimulation, there was an enhanced recruitment of NF-κB subunit p65-MTA1-pol II complex onto the -630 to -849 region of TG2 promoter [341]. Falasca et al. showed that LPS induced an enhanced NF-κB activation in macrophages in a TG2-dependent manner, therefore, upon LPS treatment, the induction of TG2 in macrophages resulted in a loop of continuous activation of NF-κB pathway. However, TG2 null macrophages treated by LPS

42 lost this activity to activate NF-κB [342]. In addition to activate NF-κB, the LPS induced TG2 in macrophages also mediated an elevated activity of phospholipase A2 (PLA2), which catalyzes the production of arachidonic acid (AA), an important substance during inflammation [343]. Sarang et al. showed that TG2 null macrophages responded to LPS treatment with increased production of IL-6 and tumor necrosis factor alpha (TNFα) due to an elevated integrin αvβ3-mediated Src kinase activity [344]. However, Yoo et al. demonstrated that TG2 null mice treated with LPS displayed reduced levels of IFN-γ and IL-6 in the serum [345]. This contradiction may arise from the discrepancy between in vivo and in vitro studies. Intriguingly, in addition to the upregulation of TG2 by pro-inflammatory LPS, anti-inflammatory mediator IL-4, also the M2 macrophage inducer, has been shown to upregulate TG2 expression in macrophages as well [346-348]. However, the IL-4 induced expression of TG2 was blunted in estrogen receptor α (ERα) deficient macrophages due to the presence of a full consensus estrogen response element (ERE) in the TG2 promoter [348]. Moreover, IL-6 is also an inducer of TG2 expression in macrophages [349]. Additionally, transforming growth factor β1 (TGF- β1) induced TG2 expression in macrophages, and TG2 was also found to promote TGF-β1 expression [350,351]. Notably, there were several studies showing contradictory results on the TG expression during macrophage differentiation. Two studies revealed that TG2 increased and FXIII-A decreased during macrophage differentiation [352,353]. Consistently, there was an upregulation of cell surface TG2 associated with β1 and β3 integrins, while no association of FXIII-A with integrins was detected and a simultaneous decrease of surface FXIII-A during macrophage differentiation was seen [323]. However, on the contrary, another three studies confirmed the increase of FXIII-A during macrophage differentiation [330,354,355]. This contradiction may arise from the use of different macrophage models in vitro. In addition, Adany et al. showed that there was a transient intranuclear accumulation of FXIII-A in the early phase of macrophage differentiation and FXIII-A in the nuclei also exhibited TG activity [356]. Studies also showed that FXIII-A was strongly induced in macrophages by the combination of IL-4 and dexamethasone, but not by either of them alone [346]. Our work demonstrated, for the first time, that TG1 was also expressed in macrophages and M-CSF strongly upregulated its expression [288,291]. Taken together, TG2, FXIII-A, and TG1 are all present in macrophages and their expressions are upregulated during macrophage differentiation. TG2 can be induced by a plethora of factors, such as pro-inflammatory LPS and anti-inflammatory IL-4, suggesting TG2 is involved in both induction and resolution stage of inflammation. It can also be induced by factors with both anti-

43 and pro-inflammatory properties such as IL-6 and TGF-β1 [357,358], indicating that TG2 may exert distinct functions in response to the same factor under various circumstances.

Transglutaminases in Macrophage Phagocytosis Phagocytosis is restricted to specialized cells, including professional phagocytes (neutrophils, monocytes, macrophages, and dendritic cells) or non-professional phagocytes (fibroblasts, endothelial, and epithelial cells). It is a receptor-mediated and actin-polymerization-dependent process for internalizing particles greater than 0.5 µm in diameter [359]. The target particles include the apoptotic cells, necrotic cell debris, and opsonized pathogens [360]. The engulfment of apoptotic cells by macrophages, also termed as efferocytosis, is crucial for tissue homeostasis, resolution of inflammation, and embryologic development [361]. The involvement of TG2 in macrophage phagocytosis was initially reported in 1981 [362], and was further confirmed by various in vitro studies [363,364]. Subsequently, studies showed that TG2 null mice displayed defective clearance of apoptotic thymocytes by macrophages in the thymus [248] and impaired clearance of apoptotic hepatocytes by Kupffer cells in the liver [247,248]. However, TG2 null macrophages exhibited unimpaired ability to ingest bacteria, yeast, opsonized non-apoptotic thymocytes [248], monosodium urate crystals [365], or oxidized low-density lipoproteins [366]. Additionally, TG2 null mice showed inflammatory infiltrates at the apoptosis sites in short term [247,248] and then developed autoimmunity in the long term [248]. Two other studies confirmed that TG2 was required for the engulfment of apoptotic cells by macrophages, but not for their recognition and binding [247,365]. These studies also confirmed that TGF-β, a cytokine which is released by macrophages engulfing apoptotic cells [367] and whose activation requires TG2 [368], was involved in the defective phagocytosis in TG2 null macrophages [247,248,365] and exogenous active TGF-β1 was able to rescue this phagocytosis defect [365]. Subsequent studies revealed that extracellular or cell surface localization of TG2 was responsible for the TG2-mediating phagocytosis and that its cross-linking function was not required for this since exogenous recombinant TG2, both wild type and catalytically inactive forms, was able to rescue the defective phagocytosis in TG2 null macrophages. Moreover, the function of TG2 in phagocytosis is dependent on its guanine nucleotide-binding pocket [365,369]. In addition, Toth et al. showed that TG2 formed a complex with milk fat globulin EGF factor 8 (MFG-E8) and integrin β3 on macrophage surface and was required for the formation of engulfing portals. In addition, there was a compensatory increase in integrin β3 expression in the TG2 null macrophages that partially corrected the

44 impaired integrin β3 signaling caused by the absence of TG2 [369,370]. In contrast to the previous findings, studies showed that TG2 cross-linking activity on the cell surface was critical for macrophage recruitment to, binding and removal of apoptotic cells, and suggested that TG2, in association with syndecan-4 on cell surface, promoted clearance of apoptotic cells by cross-linking CD44 [371]. More recently, studies unveiled a role of TG2 in regulating levels of phagocytosis-related molecules CD14 and class A macrophage scavenger receptor type I (SR-AI) in macrophages [353]. However, upregulation of TG2 alone in macrophages is not sufficient to promote their capacity to remove apoptotic cells [372]. Intriguingly, studies showed that both receptors for the Fc region of IgG (FcγR)- and complement receptor (CR)- mediated phagocytosis were strongly diminished in monocytes of FXIII-A deficient patients [354]. In addition, a human myelomonocytic cell line, negative for FXIII-A and incapable of phagocytosis, displayed FcγR- and CR-mediated phagocytosis and FXIII-A expression after differentiation to macrophages induced by TPA treatment [373]. To conclude, TG2 on the cell surface or in the extracellular matrix plays an essential role in macrophage phagocytosis of apoptotic cells via interaction with integrin β3 or other molecules/receptors. The defective clearance of apoptotic cells resulting from the absence of TG2 is associated with inflammation and autoimmune diseases. FXIII-A is involved in the FcγR- and CR-mediated phagocytosis in macrophages.

Transglutaminases in Vascular Macrophages Atherosclerosis is an immunoinflammatory vascular disorder characterized by the build-up of lipids, cholesterol, calcium, and cellular debris on the intima of medium-sized and large arteries [374]. It is well established that monocytes and macrophages play important roles in the initiation and progression of atherosclerosis [375], and that both TG2 [376,377] and FXIII-A [378] are present in human atherosclerotic plaques. Additionally, FXIII-A was found to be the dominant TG in human atherosclerotic arteries, and it may be derived from the local macrophages [379]. Studies demonstrated that irradiated low-density lipoprotein receptor (LDLR) knockout mice followed by TG2-/- bone marrow transplantation (BMT) displayed larger and deeper atherosclerotic aortic valve lesions compared with TG2+/+ BMT mice, indicating that leukocyte-derived TG2 limited the progression of atherosclerosis [366]. In this study, authors suggested that macrophage-derived TG2 may have exerted its functions by (a) promoting the phagocytosis of apoptotic cells, (b) modulating production of TGF-β which is involved in plaque stability, and (c) regulating the expression of ATP-binding cassette

45 transporter A1 (ABCA1) which mediates reverse cholesterol transport [366]. Another study showed that TG2 deficient mice exhibited a decreased collagen content and increased macrophage content in the atherosclerotic plaques, indicating a more unstable and more rupture-prone plaque [380]. In this study, FXIII-A was expressed in atherosclerotic plaques in TG2 deficient mice and suggested to compensate for the loss of TG2. FXIII-A is likely derived from macrophages and is sufficient to prevent plaque rupture by cross-linking the matrix proteins [380]. In contrast to the anti-atherogenic role of macrophage-derived TG2 and FXIII- A in later stages of atherosclerosis, studies showed that TG2 null mice displayed significantly less monocytes/macrophages on the carotid artery when exposed to oscillatory shear stress [381]. Additionally, FXIII-A inhibition by expression of the FXIII-A inhibitor FXIII-AN73- D98 decreased monocyte/macrophage infiltration of the aorta, leading to inhibition of atherosclerosis development in hypercholesterolemic apolipoprotein E deficient mice [325]. Furthermore, systemic inhibition of both TG2 and FXIII-A with the TG inhibitor L682777 did not alter the lesion size, but resulted in reduced macrophage content in the media of vessels, suggesting that TG2 and FXIII-A contribute to the early development of plaque formation via directly affecting monocyte/macrophage infiltration [382]. In conclusion, macrophage-derived TG2 and FXIII-A play an anti-atherogenic role in later stages of atherosclerosis, whereas TG2 and FXIII-A show pro-atherogenic function in the early stages of atherosclerosis by regulating monocyte/macrophage infiltration.

Transglutaminases in Alveolar and Interstitial Macrophages In the lung, there are two major populations of macrophages present, alveolar and interstitial macrophages. Alveolar macrophages reside within the lumen of alveolus, whereas interstitial macrophages locate to the interalveolar spaces [383]. Studies showed that cigarette smoke inactivated TG2 by attacking the active-site cysteine residue via an oxidative mechanism in alveolar macrophages [384,385]. In addition, there was a decrease in TG2 activity in the alveolar macrophages of smokers compared to nonsmokers [386]. Another study showed that TG2 inhibition decreased the production of cysteinyl leukotriene (a product derived from AA) in IL-4-treated macrophages, suggesting a role for TG2 in airway inflammation [387]. FXIII- A is also present in alveolar macrophages [388]. The levels of FXIII-A derived from activated or injured alveolar macrophages were elevated in the bronchoalveolar lavage fluid from patients with chronic bronchoalveolar inflammation [389]. The role of TGs in interstitial macrophages is an interesting topic but there is a paucity of information. The limited available

46 literature showed that lung interstitial macrophages from the ovalbumin-induced asthmatic mice were polarized towards the M2 phenotype with an increased expression of TG2, which is a marker of M2 macrophages [390]. These findings suggest that macrophage-derived TG2 and FXIII-A play a role in airway inflammation, but further research is required to determine their accurate functions.

Transglutaminases in Microglia Microglia, resident macrophages in the central nervous system (CNS), represent 5%-20% of the total glial cell population [391]. Microglia have multiple functions and are involved in various pathological processes, such as neuroinflammation, neurodegeneration, ischemia, and trauma [392]. Microglia are usually maintained in a resting or relatively inactive state with a “ramified” morphology with many short and fine processes. These processes undergo continuous cycles of extension and retraction to scan their local environment [393]. In response to trauma, infection, or infarction, microglia are activated and transform into “amoeboid” morphology (i.e., spherical shape without processes) and display phagocytosis [393]. These cells are morphologically indistinguishable from infiltrated monocytes/macrophages in the CNS. Alzheimer’s disease (AD) is characterized by neuronal loss and accumulation of pathogenic amyloid-β (Aβ) assemblies in the brain, and this Aβ-induced neuronal loss was caused by the enhanced microglial phagocytosis [394]. In addition, increased NO production of microglia also resulted in neuronal loss in neurodegenerative diseases [395]. Studies showed that LPS increased TG2 expression in BV-2 microglia. This induction of TG2 promoted the NO production and enhanced the phagocytosis of dead cells in microglia [396,397] via the TG2- mediated activation of NF-κB pathway [398]. In addition, amphotericin B upregulated the expression of TG2 and inducible NO synthase (iNOS) and increased the phagocytosis of dead cells in BV-2 microglia [399]. Recent studies showed that the uptake of Aβ by microglia was mediated by forming a complex of aggregated Aβ/MFG-E8/TG2 [400]. In order to explore the roles of TG2 in the pathogenesis of AD, many studies were carried out using THP-1 cells, a monocyte-like cell line. Studies showed that Aβ treatment increased TG2 expression in THP- 1 cells and TG2 was required for the Aβ-induced monocyte maturation and activation [401]. Another study confirmed that TG2 mediated the Aβ-induced THP-1 monocyte activation via AP1/JNK signaling pathways [402]. Additionally, FXIII-A positive microglia were found abundant in primitive plaques in parietal cortex of AD brains, suggesting FXIII-A plays a role

47 in the early stage of AD pathology [403]. Multiple sclerosis (MS) is an inflammatory demyelinating CNS disease, characterized by sensory, motor, and cognitive deficits [404,405]. The principal study model for MS is the experimental autoimmune encephalomyelitis (EAE) [406]. In marmosets with EAE, TG2 was expressed by infiltrating monocytes in active white matter lesions. In addition, TG2 was co- localized with integrin β1 and closely associated with extracellular fibronectin, suggesting that TG2 plays a crucial role in the adhesion and migration of infiltrating monocytes in EAE. In contrast to the white matter lesions, TG2 was expressed mainly by resident microglia in grey matter lesions. However, fibronectin expression was absent suggesting an alternative role for microglia-derived TG2 in grey matter lesions [407]. Studies also confirmed the presence of TG2 in infiltrated macrophages in human MS lesions [408]. Furthermore, in rats with EAE, inhibition of TG2 activity resulted in clinical improvement and attenuated demyelination, indicating a role for TG2 during MS pathogenesis [408]. To determine if the cross-linking activity of TG2 caused these effects, TG2 inhibitors, BJJF078 and ERW1041E (neither of them interferes with TG2 binding to fibronectin), were administered in this model. Studies showed that only BJJF078 inhibited cellular TG2 activity in THP-1 cells and only ERW1041E resulted in attenuated EAE disease motor-symptoms, suggesting that extracellular TG2 activity, rather than cellular activity, is more likely involved in mouse EAE pathology [409]. In a newly developed mouse model which enables the in vivo visualization of monocytes during EAE, TG2 was present in monocytes in the spinal cord lesions and TG2 positive monocytes attached to the endothelial lumen of the blood vessel, suggesting that TG2 may contribute to the extravasation of monocytes into the CNS during EAE [410]. Intriguingly, recent studies revealed that microglia-derived TG2 promoted myelin formation and repair by oligodendrocytes [411]. Microglia-derived TG2 was also associated with neuronal death induced by ischemia/reperfusion [412]. In addition, studies showed that oligomerization of superoxide dismutase 1 catalyzed by TG2 induced activation of BV-2 microglia, accelerating neuroinflammation in amyotrophic lateral sclerosis (ALS) [413]. Furthermore, studies showed hyperhomocysteinemia was associated with various chronic neurodegenerative diseases, such as AD, MS, and ALS, and TG2 played an essential role in homocysteine-induced activation of THP-1 monocytes [414]. Collectively, these findings reveal a contributing role of TG2 to the pathogenesis of various neurological disorders such as AD, MS, and ALS via regulating phagocytosis and activation of microglia and affecting monocyte infiltration into the CNS.

Transglutaminases in Osteoclasts Bone is a mineralized form of connective tissue and its integrity is maintained by constant remodeling through the balanced activities of two cell types: the bone-forming osteoblast and the bone-degrading osteoclast, a specialized bone-resident multinucleated macrophage [415]. Starting from monocytes/macrophages, osteoclast differentiation requires two essential cytokines, M-CSF and receptor activator of nuclear factor-kappaB ligand (RANKL). Osteoclastogenesis is a complex process involving a differentiation stage, followed by a cell fusion and multinucleation stage [16]. When osteoclasts attach to bone, they generate two polarized structures: the bone-apposed ruffled border and the sealing zone which seals off a resorption compartment between the cell membrane and the bone surface. Protons are pumped and proteolytic enzymes are secreted from the ruffled border side of the osteoclast into the resorption compartment to dissolve minerals and degrade bone matrix proteins [46]. Increased osteoclast activity is responsible for bone destruction in diseases such as osteoporosis, periodontitis, and rheumatoid arthritis [416]. Recently, four studies were published by us and others exploring the role of TGs in osteoclastogenesis and bone resorption. Raghu et al. showed that FXIII-A deficient mice displayed reduced osteoclastogenesis in vivo and in vitro and that TG inhibitor, cystamine, suppressed osteoclastogenesis in vivo [289]. Furthermore, our work showed that TG2 and FXIII-A double knockout mice exhibited severe osteopenia due to increased bone resorption. In addition, the double FXIII-A/TG2 deficient macrophages gave rise to increased osteoclastogenesis in vitro, suggesting that the two enzymes negatively regulate osteoclastogenesis [288]. However, a non-specific TG inhibitor, NC9, blocked osteoclastogenesis in this study, leading to the discovery of TG1 being expressed in wild type and FXIII-A/TG2 deficient macrophages and osteoclasts [288]. Subsequently, Kim et al. showed that suppressed expression of TG2 by siRNA resulted in increased osteoclastogenesis in vitro and that TG2 deficient mice displayed increased osteoclast number and lower trabecular bone mass in vivo [290]. However, we did not see this phenotype in TG2 knockout mice [288]. Most recently, our work showed that NC9 inhibited differentiation, migration, and fusion of pre-osteoclasts, as well as resorption of mature osteoclasts. NC9 also increased RhoA levels and blocked podosome belt formation, suggesting TG activity regulates actin dynamics in osteoclasts. Consistently, this study showed that TG1, TG2 and FXIII-A co-localized to podosomes in osteoclasts [291]. Furthermore, studies showed that Z006 (a TG inhibitor) became cell permeable at 40 µM [228], and Z006 only exhibited inhibitory effect on osteoclast

49 differentiation at 40 µM or higher concentrations, indicating that intracellular TG activity is involved in osteoclastogenesis [291]. Another two interesting studies revealed an indirect role of TG2 in regulating osteoclast differentiation. One study showed that TG2 inhibition resulted in reduced RANKL expression in macrophages and human periodontal ligament cells [417]. Another study showed that TG2 mediated 1α,25-dihydroxyvitamin D3-induced macrophage fusion in a spermidine-dependent mechanism [418]. Although these findings may seem contradictory, they are not incompatible. Since three TGs are present in osteoclasts whose differentiation involves several cellular events, it is possible that they all have separate functions, and/or may be up- or downregulated in the absence of others, which would complicate interpretation of null mouse models. Future work should be directed to dissect out their separate functions and joint effects on osteoclasts.

Transglutaminases in Other Macrophages Studies showed that pro-inflammatory LPS and IFN-γ induced secretion of thioredoxin-1 (TRX) by macrophages was able to activate the extracellular TG2 in macrophages in vitro and that intravenous administration of TRX resulted in a rapid increase in TG2 activity exclusively in the small intestine in mice (i.e., this TRX-induced TG2 activity was not observed in other organs), suggesting that enhanced TG2 activity may be caused by inflammation in celiac disease [419]. Previous studies showed that cross-linking activity of TG2 was required for apoptotic envelope formation on macrophages infected with attenuated Mycobacterium tuberculosis (MTB) strain and thus the apoptotic macrophages limited bacterial replication and dissemination [420]. Recent studies showed that both inactivation and inhibition of TG2 suppressed the intracellular replication of virulent MTB in macrophages via impairing cell autophagy [421]. Another two interesting studies showed that cysteamine, a TG inhibitor (albeit not specific) [422], improved the autophagy of macrophages derived from patients or mice with cystic fibrosis, resulting in enhanced bacterial clearance by macrophages [423,424]. Studies also demonstrated that peritoneal injection of cystamine reduced the activity of matrix metalloproteinase-9, the expression of TNF-α and TGF-β in macrophages, and suppressed the production of anti-cardiolipin autoantibody in mice with systemic lupus erythematosus [425].

2.7 Conclusions Macrophages are highly heterogeneous immune cells found in multiple tissues. Monocytes, previously considered as the immediate upstream precursors of macrophages, infiltrate into

50 tissues upon damage or infection. Depending on the tissue context, they exert distinct functions, which can considerably overlap with tissue-resident macrophages. Over the last 30 years, numerous studies suggest that TGs, particularly TG2 and FXIII-A, participate in monocyte- and macrophage-mediated physiological and pathological processes. Firstly, TG2 and FXIII-A contribute to monocyte adhesion and extravasation, a process essential for the initiation of inflammation, especially for the pathogenesis of atherosclerosis and MS. Next, TG2, FXIII-A, and TG1 are all present in macrophages and their expressions are upregulated during macrophage differentiation. TG2 can be induced by both pro-inflammatory and anti- inflammatory substances, suggesting TG2 is involved in both induction and resolution stage of inflammation. Among these mediators, LPS is the most important one due to its association with septic shock [342,345]. Furthermore, TG2 and FXIII-A play critical roles in different types of macrophage phagocytosis. TG2 on the macrophage surface or in the extracellular matrix interacts with integrin β3 or other molecules/receptors to mediate phagocytosis of apoptotic cells, a process crucial for resolution of inflammation and for prevention of development of autoimmune diseases. FXIII-A is involved in FcγR- and CR-mediated phagocytosis in macrophages. Finally, we reviewed the roles of TGs in tissue-resident macrophages and infiltrated monocytes in different tissue environments. In vasculature, macrophage-derived TG2 and FXIII-A play an anti-atherogenic role in later stages of atherosclerosis, whereas TG2 and FXIII-A show pro-atherogenic function in the early stages of atherosclerosis by regulating monocyte/macrophage infiltration. In the lung, macrophage-derived TG2 and FXIII-A play a role in airway inflammation, but further research is required to accurately determine their functions. In CNS, TG2 participates in the pathogenesis of various neurological disorders such as AD, MS, and ALS via regulating phagocytosis and activation of microglia and affecting monocyte infiltration. In bone, TG2, FXIII-A, and TG1 are all present in osteoclasts. Although different mouse models show contradictory results regarding the role of TGs on osteoclast differentiation, in vitro studies show TG activity is required for osteoclastogenesis. Future work should be directed to dissect out their separate functions and joint effects on osteoclasts. However, the mechanisms of action of TGs in these processes are not yet fully understood. Thus, further research is needed to determine the extracellular or intracellular localization of TGs, the catalytic or non-catalytic activities involved, and target substrates in monocytes and macrophages. This will benefit from many advancements in currently available TG research tools, including TG-specific inhibitors [226-228,426-428] and substrate peptides [211-216].

2.8 Acknowledgments Funding: This work was supported by grants to M.T.K. from the Canadian Institutes of Health Research (CIHR) (MOP-119403).

Chapter 3 - Transglutaminase activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics

Our research group has shown that three TGs are present in osteoclasts and their precursors and Tgm2-/-;F13a1-/- monocytes display increased potential to form osteoclasts in vivo and in vitro. However, chemical inhibition of TG activity with NC9 completely blocks osteoclast formation in both wild type and TG2/FXIII-A deficient monocytes, suggesting TG activity is required for osteoclastogenesis. However, the exact stages of osteoclast formation affected by NC9 and its mechanism of action have remained unknown. In this study, I have investigate the role of TG activity in different stages of osteoclastogenesis in vitro using NC9. My results show that NC9 inhibits differentiation, migration, and fusion of pre-osteoclasts as well as the podosome belt formation and resorption of osteoclasts. In addition, NC9 increases RhoA levels and the inhibitory effect of NC9 on osteoclastogenesis and podosome belt formation can be completely reversed with a Rho-family inhibitor Exoenzyme C3, suggesting that TG activity regulates actin dynamics in osteoclasts. This is the first report demonstrating that TG activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics.

The study presented in this chapter has been published in the peer-reviewed journal Journal of Cellular Physiology, Vol 233, Issue 9, p. 7497-7513, 2018  Wiley Online Library

Transglutaminase activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics

Sun H1 and Kaartinen MT1,2 1 Division of Biomedical Sciences, Faculty of Dentistry, McGill University, Montreal, QC, Canada 2 Division of Experimental Medicine, Department of Medicine, Faculty of Medicine, McGill University, Montreal, QC, Canada

3.1 Abstract Osteoclasts, bone resorbing cells, derive from monocyte/macrophage cell lineage. Increased osteoclast activity is responsible for bone destruction in diseases such as osteoporosis, periodontitis and rheumatoid arthritis. Transglutaminases (TGs), protein cross-linking enzymes, were recently found involved in osteoclastogenesis in vivo, however their mechanisms of action have remained unknown. In this study, we have investigated the role of TG activity in osteoclastogenesis in vitro using four TG inhibitors, NC9, Z006, T101, and monodansyl cadaverine. Our results showed that all TG inhibitors were capable of blocking the entire osteoclastogenesis process. The most potent of the inhibitors, NC9 when added to cultures at different phases of osteoclastogenesis, inhibited differentiation, migration, and fusion of pre- osteoclasts as well as resorption activity of mature osteoclasts. Further investigation into the mechanisms revealed that NC9 increased RhoA levels and blocked podosome belt formation suggesting that TG activity regulates actin dynamics in pre-osteoclasts. The inhibitory effect of NC9 on osteoclastogenesis as well as podosome belt formation was completely reversed with a Rho-family inhibitor Exoenzyme C3. Microtubule architecture, acetylation, and detyrosination of α-tubulin were not affected. Finally, we demonstrated that macrophages and osteoclasts expressed mRNA of three TGs: TG1, TG2, and Factor XIII-A which were all differentially regulated in these cells during differentiation. Immunofluoresence microscopic analysis showed that all three enzymes co-localized to podosomes in osteoclasts. Taken together, our data suggests that TG activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics and that this may involve contribution from all three TG enzymes.

3.2 Introduction Bone resorption is orchestrated by osteoclasts which are large multinucleated cells derived from monocyte/macrophage lineage. The differentiation from monocytes/macrophages to mature osteoclasts is an intricate process involving a differentiation stage which is followed by a cell fusion and multinucleation stage. These complex events require both extensive interactions between cytokines and receptors, and dynamic rearrangement of cytoskeleton and plasma membrane [16]. Osteoclast differentiation requires two essential cytokines, macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor-κB ligand (RANKL) [429]. M-CSF mediates proliferation, differentiation and survival of osteoclast precursors and promotes expression of RANK in osteoclast precursors, priming the RANK-positive cells to respond to RANKL [430] which in turn induces sequential expression of NF-κB, c-Fos and NFATc1 [85]. NFATc1 is the master regulator of osteoclast differentiation and regulates osteoclast-specific genes such as tartrate resistant acid phosphatase (TRAP) [120] and cathepsin K (CTSK) [431] as well as matrix metalloproteinase-9 (MMP-9) [432]. With the continuing stimulation of RANKL, mononuclear TRAP-positive (TRAP+) pre-osteoclasts migrate toward each other and fuse to form multinucleated osteoclasts. This fusion process is regulated by two transmembrane proteins-dendritic cell-specific transmembrane protein (DC- STAMP) and osteoclast-stimulatory transmembrane protein (OC-STAMP) [135,137,139,433], both transcriptionally regulated by NFATc1 [121,138]. Unlike most non-transformed cells relying on F-actin stress fibers or focal adhesion plaques, cells from monocytic lineage such as dendritic cells, macrophages and osteoclasts develop podosomes to adhere and migrate with [434]. Podosomes are highly dynamic dot-like structures, containing an F-actin core and a surrounding F-actin cloud, along with a variety of molecules such as integrins, kinases and small GTPases (Rho and Rac) [147]. On non-mineralized substrates, podosomes in osteoclasts can self-assemble into clusters, rings and finally form a stable peripheral actin belt [147]. The assembly/disassembly of podosome superstructures allows osteoclast migration during fusion process and during bone resorption [142,435]. Transglutaminases (TGs; EC 2.3.2.13) are a family of thiol- and Ca2+-dependent acyl transferases that catalyze the formation of a covalent bond between γ-carboxamide groups of peptide-bound glutamines and free amine groups (e.g., protein- or peptide-bound lysines) [188,190,253]. Nine TG genes are present in humans, TG1-7 and Factor XIII-A (FXIII-A), which are catalytically active enzymes. One family member, erythrocyte membrane protein band 4.2, is inactive as a transglutaminase [190]. So far, TG2 and Factor XIIIA have been

55 shown involved in cell adhesion, migration [323], differentiation [356], and in function of monocytes and macrophages [295,296,436]. TGs have been long reported expressed and found in bone and osteoblasts [280,281,284-286,437-441]. Three recent studies by us and others suggest their role also in osteoclastogenesis and bone resorption [288-290]. The work of Raghu et al. showed that FXIII-A deficiency in mice led to reduced osteoclastogenesis in vivo and in vitro and that transglutaminase inhibitor, cystamine inhibited osteoclastogenesis in vivo [289]. Furthermore, our recent work showed that deletion of both TG2 and FXIII-A gene expression in a double knockout mouse model resulted in severe osteopenia caused by increased bone resorption [288]. The increased osteoclastogenesis was also seen in double FXIII-A/TG2 deficient monocytes in vitro demonstrating that the two enzymes negatively regulate osteoclastogenesis. However, similarly, as in the work of Raghu et al., a chemical inhibitor of TGs blocked osteoclastogenesis in our study, which was attributed to expression of TG1 in wild type and FXIII-A/TG2 deficient monocytes and osteoclasts [288]. Although we did not see a bone phenotype in either, individual TG2 or FXIII-A knockouts [288], Kim et al. showed recently that TG2 deficient mice exhibited increased osteoclast number and lower trabecular bone mass in vivo and that siRNA knockdown of TG2 gave rise to increased osteoclastogenesis in vitro [290]. The data from these studies strongly suggest that TGs are an important part of osteoclastogenic program and suggest an interplay between the three TGs in the process. The aim of this study was to investigate effect of TG inhibitors on osteoclastogenesis and to shed light on the mechanisms how they may exert their effect. We report that there are indeed three TGs expressed in macrophages and osteoclasts and that TG activity regulates differentiation, as well as migration and fusion of osteoclasts via affecting actin dynamics in these cells.

3.3 Materials and methods Reagents and antibodies MEM Alpha (αMEM) (12561-056), penicillin-streptomycin, L-glutamine, sodium pyruvate were from Gibco (Burlington, ON, Canada). Fetal bovine serum was from Hyclone (Waltham, MA). Human M-CSF and human sRANK Ligand were from PeproTech (Rocky Hill, NJ). NC9 was synthesized by Gene Tech Inc (Indianapolis, IN) [286]. Monodansyl cadaverine (MDC) was from Sigma-Aldrich (St Louis, MO). Z006 and T101 were from Zedira GmbH (Darmstadt, Germany). RhoA inhibitor (Exoenzyme C3) (CT04) was from Cytoskeleton (Denver, CO). All other reagents unless otherwise specified were purchased from Sigma-Aldrich (Oakville, ON, Canada) or Fisher Scientific (Hampton, NH).

Mouse monoclonal TG1 antibody (E-6) was from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse monoclonal TG2 (Ab-1) antibody was from Fisher Scientific. Rabbit anti-mouse FXIII- A (675-688 peptide sequence) (polyclonal antibody) was designed and generated by GenScript corporation (Piscataway, NJ) [286]. Mouse monoclonal acetylated α-tubulin antibody (6-11B- 1), rabbit polyclonal detyrosinated α-tubulin antibody were from Abcam (Cambridge, MA). Rabbit polyclonal actin antibody, mouse monoclonal α-tubulin antibody (DM1A), mouse monoclonal biotin antibody (BN-34) were from Sigma-Aldrich. Rabbit polyclonal fibronectin antibody was from EMD Millipore (Billerica, MA). Mouse IgG1 kappa isotype control, mouse IgG2a kappa isotype control, Rabbit IgG isotype control, secondary antibodies Alexa Fluor® 568 and 488 conjugates, Alexa Fluor® 488-phalloidin and DAPI (4', 6-diamidino-2- phenylindole) were from Thermo Fisher Scientific (Rockford, IL). Secondary antibody HRP- linked anti-mouse IgG was from GE Healthcare (Mississauga, ON, Canada). HRP-linked anti- rabbit IgG was from Cell Signaling (Whitby, ON, Canada).

Osteoclast differentiation from mouse bone marrow cells Studies were compliant with McGill University guidelines established by the Canadian Council on Animal Care. Mouse bone marrow cells were collected from 6 to 10 weeks old C57BL/6 mice (from Jackson Laboratories) as described previously with minor modifications [442]. Briefly, bone marrow cells were plated at a density of 15  106 cells per 75 cm2 tissue culture flask and cultured overnight in αMEM supplemented with 10% FBS, 1% penicillin- streptomycin, 1% L-glutamine solution, 1% sodium pyruvate and 25 ng/ml M-CSF. After 24 hr, non-adherent cells were collected and plated at 5  104 cells/cm2 and incubated with 50 ng/ml M-CSF. After 48 hr, the non-adherent cells were washed out and adherent cells were considered to be M-CSF-dependent bone marrow-derived macrophages and used as osteoclast precursors. Osteoclast differentiation was then induced by 50 ng/ml M-CSF and 50 ng/ml RANKL for 5 or 6 days with two medium changes on day 2 and 4. The bone marrow macrophages were also treated with only 50 ng/ml M-CSF for 5 or 6 days to serve as negative control cultures. Only joint treatment with M-CSF and RANKL results in osteoclastogenenesis. TG activity inhibitor NC9, MDC and Z006 were dissolved in dimethylsulfoxide (DMSO), T101 in PBS. RhoA inhibitor (Exoenzyme C3) was dissolved in sterile water.

Cell viability assay Cell viability in the presence of inhibitors was determined using the methyl-thiazol tetrazolium

(MTT) assay. Bone marrow macrophages were plated in 96-well plates and incubated with M- CSF and RANKL in the presence or absence of NC9, MDC, Z006, and T101. After 5 days, 0.25 mg/ml MTT reagent was added and incubated for 3 hr. Purple precipitates were dissolved in DMSO. The absorption was measured at 560 nm using a microplate reader (TECAN infinite F200 PRO, Männedorf, Switzerland). The experiments were independently repeated three times, each time in triplicate.

TRAP staining and cell counting Bone marrow macrophages were plated in 96-well plates and incubated with M-CSF and RANKL in the presence or absence of different concentrations of NC9, MDC, T101, Z006 and 5 ng/ml Exoenzyme C3. Controls were treated with the drug vehicles. At the end points, cells were fixed with 3.7% formaldehyde and stained for TRAP with the TRAP Staining Kit (Cosmo Bio, Carlsbad, CA). Pre-osteoclasts were identified as mononuclear TRAP+ cells and osteoclasts were identified as multinucleated (≥ 3 nuclei) TRAP+ cells. Cells were then stained with DAPI to facilitate counting nuclei. The experiments were independently repeated three times, each time in triplicate. For each experimental condition, five non-overlapping images per well were taken under 10  objective to evaluate osteoclast differentiation.

TRAP activity assay TRAP activity in the culture supernatant was measured with the TRAP Staining Kit (Cosmo Bio) as described previously with minor modifications [443]. Briefly, chromogenic substrate was dissolved with tartrate-containing buffer and this reaction buffer was added to the culture supernatants in a 96-well plate. The plate was then incubated for 3 hr in the dark at 37 °C and then read in a microplate reader at 540 nm (Molecular Devices SpectraMax M2e, Sunnyvale, CA). The experiments were independently repeated three times, each time in triplicate.

Osteoclastic resorption assay Osteoclastic resorption assays were conducted on Corning® Osteo Assay Surface 24-well plates coated with inorganic crystalline calcium phosphate (Tewksbury, MA). Bone marrow macrophages were plated and osteoclastogenesis was induced as described above. After 5 days, mature osteoclasts were cultured for further 5 days with M-CSF and RANKL in the presence or absence of NC9. Resorption pits were visualized by light microscopy after modified von Kossa staining [444] and quantified with ImageJ (NIH). The experiments were independently

58 repeated three times, each time in triplicate. For each experimental condition, five images per well were taken under 5  objective to evaluate osteoclast resorption.

Migration assay Cell migration was measured using a cell exclusion zone assay [445]. Bone marrow macrophages were plated into Oris™ 96-well plate with Oris™ cell seeding stoppers to prevent cells from attaching to the central zone with a diameter of 2 mm (Platypus Technologies, Madison, WI). Cells were cultured with M-CSF and RANKL for 2 days and the stoppers were removed to reveal the exclusion zone into which pre-osteoclasts were then allowed to migrate with treatment of M-CSF and RANKL in the presence or absence of 20 µM NC9 for 12 hr. The exclusion zones were monitored under 5  objective and the number of migrated cells were quantified by counting cells that crossed into the void zone during the 12 hr. The experiments were independently repeated three times, each time in triplicate.

Assessment of podosome belt formation Bone marrow macrophages were differentiated for 5 days with M-CSF and RANKL in the presence or absence of 10 µM NC9 and 5 ng/ml Exoenzyme C3. When the osteoclast formed, cells were fixed with 3.7% formaldehyde and labeled with Alexa Fluor® 488-phalloidin to visualize specific F-actin structure-containing podosome belts. The extent of podosome belt formation was quantified by counting the podosome belts versus the total number of osteoclasts and expressed as percentage. The experiments were independently repeated three times, each time in triplicate. For each experimental condition, three images per well were taken under 10  objective to evaluate osteoclast podosome belt formation.

59 and individual proteins were detected with specific antibodies as well as corresponding secondary antibodies conjugated with horseradish peroxidase. Bands were visualized using the ECL Plus kit (GE Healthcare), and chemiluminescence was detected using ChemiDocTM Touch Imaging System (Bio-Rad).

In vitro TG activity assay In vitro TG activity assay was performed as described previously [437]. Briefly, cell lysate was incubated for 2 hr at 37 °C with 1 μg bovine plasma fibronectin (pFN), 2 mM 5-(biotinamido) pentylamine (bPA), in a reaction buffer containing 1 mM dithiothreitol (DTT), 3 mM CaCl2, 10 mM Tris-HCl (pH 8.0). After incubation, samples were analyzed by SDS-PAGE and western blotting as above to visualize covalent bPA incorporation to pFN which reflects the TG activity present in the samples.

RhoA activity assay GTP-bound active RhoA was quantified using the RhoA Pull-down Activation Assay Biochem Kit (Cytoskeleton) according to the manufacturer's protocol. Briefly, bone marrow macrophages were cultured with M-CSF and RANKL in the presence or absence of 20 µM NC9 for 2 days. After the cell lysate collection and protein concentration measurement, GTP- bound RhoA was pulled down from whole-cell extracts using Rhotekin RBD beads. Affinity precipitates and whole-cell lysates were analyzed by SDS-PAGE and western blotting as above. The experiments were independently repeated three times and protein bands were quantified with ImageJ (NIH).

Immunofluorescence microscopy Bone marrow macrophages were plated on a 96-well plate and cultured as described above. On day 5, cells were fixed with 3.7% formaldehyde and blocked with 2% bovine serum albumin (BSA). This was followed by overnight incubation with primary antibodies at 4 °C, and a washing step with 0.1% BSA and incubation with Alexa Fluor® conjugated secondary antibodies. F-actin was labeled with Alexa Fluor® 488-phalloidin and nuclei were stained by DAPI. Images were taken under 20  objective by using the Leica DMi8 inverted fluorescence microscope and Leica Application Suite X software (version 3) (Leica, Concord, ON, Canada). Controls included elimination of primary antibodies from experiments and isotype controls for each antibody used.

RNA extraction, RT-PCR and qPCR Total RNA from cell cultures was extracted using the RNeasy Mini Kit (Qiagen, Venlo, Netherlands), followed by the DNA Removal Kit (Thermo Fisher Scientific) and then quantified by the TECAN infinite F200 PRO with the NanoQuant Plate. The RT-PCR was performed with One-Step RT-PCR Kit (Applied Biological Materials, Richmond, BC, Canada) on the T100 Thermal Cycler (Bio-Rad). PCR products were analyzed by 2% agarose gel electrophoresis. Bands were detected using ChemiDocTM Touch Imaging System (Bio-Rad). Primers used were previously described [281]. The experiments were independently repeated twice. For qPCR, cDNA was synthesized from RNA with the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). Real-time PCR was performed on the StepOnePlus Real-Time PCR System (Applied Biosystems). TaqMan® Fast Advanced Master Mix and primers were purchased from Applied Biosystems. Expression levels of Tgm1 (Mm00498375_m1), Tgm2 (Mm00436987_m1), F13a1 (Mm00472334_m1), Nfatc1 (Mm00479445_m1), Trap/Acp5 (Mm00475698_m1), Ctsk (Mm00484039_m1), Mmp9 (Mm00442991_m1), Dcstamp (Mm04209234_m1), and Ocstamp (Mm00512445_m1) were normalized to Gapdh (Mm99999915_g1). The experiments were independently repeated three times, each time in triplicate.

Statistical analysis Data were analyzed with GraphPad Prism software (version 5.0). Results were presented as  SEM (standard error of the mean) of three independent experiments done in triplicates. p values < 0.05 were considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001).

3.4 Results Inhibition of TG activity with NC9 blocks osteoclastogenesis To begin our investigation whether and how TG activity regulates osteoclast differentiation, bone marrow macrophages were cultured with M-CSF and RANKL for 5 days in the absence or presence of NC9, which is a peptidic, irreversible TG inhibitor bearing a warhead moiety that attacks and blocks the enzyme active site by binding to it covalently [286]. RANKL treatment induced the formation of TRAP+ osteoclasts as expected, and this induction was dose-dependently and significantly inhibited by NC9 at 5-50 µM concentrations (Figures 1a and 1b). Counting the number of TRAP+ osteoclasts (nuclei ≥ 3) showed an initial, significant

61 increase in osteoclast number which was followed by a drastic decrease and then a complete blockage of osteoclast formation (Figure 1c). Quantification of osteoclast size by counting the nuclei per osteoclast demonstrated that NC9 reduced the cell size in a dose-dependent manner explaining the sudden increase in numbers of osteoclasts at lower concentrations, suggesting that TG activity inhibited osteoclast fusion (Figure 1d). To exclude the possibility that NC9 is toxic to macrophages/osteoclasts, cell survival analysis with MTT assay was performed. As shown in Supplementary Figure S1a, cells tolerated NC9 up to 50-60 µM level and no decrease in cell numbers was seen. The apparent increase in cell numbers is likely due to proliferation of macrophages which was permitted once osteoclastogenesis was inhibited.

NC9 and other TG inhibitors T101, Z006, and MDC inhibit both differentiation and fusion of pre-osteoclasts Osteoclastogenesis can be divided into two steps. Monocyte/macrophage lineage precursor cells first differentiate into TRAP+ mononuclear pre-osteoclasts which is followed by their fusion and further differentiation into multinucleated osteoclasts (Figure 2a). In our 5-day culture system, these two stages correspond to the first 2 days and the last 3 days respectively (Supplementary Figure S2). Thus, to examine at which stage NC9 suppresses osteoclastogenesis, cells were exposed to NC9 either for the first 2 days (Exp 1) or for the last 3 days (Exp 2) (Figure 2a). As shown in Figures 2b and 2c, the number of TRAP+ mononuclear pre-osteoclasts was significantly decreased by NC9 treatment for the first 2 days. Furthermore, the size of multinucleated osteoclasts was significantly reduced by NC9 for the last 3 days. Taken together, our results suggested that, indeed, NC9 inhibited both differentiation and fusion of pre-osteoclasts. This was confirmed via analysis of the expression of the major osteoclast differentiation markers (Nfatc1, Trap, Ctsk, Mmp-9) and fusion markers (Dcstamp, Ocstamp) under same experimental conditions (Exp 1 and Exp 2) in Figure 2a. Data presented in Figure 3a-f showed that NC9 in Exp 1 strongly reduced the expression of all the differentiation and fusion markers, while NC9 in Exp 2 only reduced the expression of Dcstamp. To further provide evidence of TG involvement in osteoclastogenesis, we examined whether other types of TG activity inhibitors could give rise to the same results. Monodansyl cadaverine (MDC) is a dansyl containing primary amine that competes with lysine side chain in cross- linking reaction and integrates to proteins upon TG activity and thus blocks protein polymerization. Z006, a “DON” compound, is an irreversible and specific blocker of TG2 [446]. T101, a 2-[(2-oxopropyl)thio] imidazolium derivate, is a reversible blocker of TG2 and

FXIII-A [447,448]. As seen in Figure 4, all the three inhibitors blocked osteoclast differentiation as evidenced by complete absence of TRAP+ cells and the undetectable TRAP activity in the culture supernatant (Figure 4a-c). As evident from Supplementary Figure S1b-d, cells tolerated all these inhibitors to the maximum concentrations tested.

Inhibition of TG activity with NC9 suppresses osteoclastic resorption To measure the effect of NC9 on osteoclast function, resorption activity of mature osteoclasts was assessed. Mature osteoclasts (treated with M-CSF and RANKL for 5 days) were formed on a bone-like mineral surface and treated with M-CSF and RANKL for another 5 days in the absence or presence of NC9. During this time, osteoclasts migrate and resorb the bone-like surface which can be visualized and quantified via von Kossa staining. As shown in Figures 5a and 5b, osteoclast resorption was significantly and dose-dependently reduced by NC9. TRAP staining during osteoclastic resorption assay confirmed the presence of osteoclasts (Supplementary Figure S3).

TG activity is required for pre-osteoclast migration TGs, particularly TG2 has been linked to cell migration in previous reports [238,323,449,450]. As the migration of pre-osteoclasts (toward other cells) is necessary for their fusion as well as for bone resorption process (to move along the surface), the effect of NC9 on the migration of pre-osteoclasts was investigated using a cell exclusion zone assay. Pre-osteoclasts were formed in a well with a cell seeding stopper to generate a central exclusion zone after which cells were treated with M-CSF and RANKL in the presence or absence of 20 μM NC9 for 12 hr. Analysis of the numbers of migrated cells into the exclusion zone during this time showed that NC9 inhibited the process in a significant manner (Figures 6a and 6b), demonstrating that TG activity is important for pre-osteoclast migration. The end point of 12 hr was chosen to eliminate potential increase in cell numbers due to proliferation as cell division generally takes 24 hr [451]. Thus the cells found in exclusion zone at 12 hr time point can only arise from migration.

TG inhibitor NC9 increases RhoA levels, decreases podosome belt formation - the inhibitory effect is completely reversed with a Rho-family inhibitor As shown in Figure 1, NC9 induced a strong retraction and thickening of osteoclasts as evidenced by the darker staining of TRAP. Previous studies have shown that RhoA activation

63 promotes cell retraction whereas RhoA inhibition allows cell spreading and flattening in macrophage-derived multinucleated giant cells [157]. Therefore, RhoA activity was investigated in NC9 treated osteoclasts. As shown in Figure 7a-c upon inhibition with NC9 for 2 days, pre-osteoclasts showed significantly increased total RhoA levels as well elevated (albeit not significant) RhoA activity compared to the control group, suggesting that NC9 may mediate its effects via RhoA in osteoclasts. It has been shown that podosome belt forms under low RhoA activity, that is, podosome belt formation is accelerated when RhoA is inhibited [159]. To examine this, cells were stained for actin to visualize podosome belts. As seen in Supplementary Figure S4a, cells lost their podosome belts at the periphery when treated with NC9 and only actin clusters were found. To gain further evidence that increased RhoA level/activity is responsible for NC9 effects, we attempted to rescue NC9-mediated osteoclastogenesis defect with a Rho-family inhibitor, Exoenzyme C3. Adding 5 ng/ml Exoenzyme C3 to 10 µM NC9 treatment normalized the number and size of osteoclasts as well as rescued the podosome belt formation defect (Figures 7d and 7e). In normal, control cultures, 75% of the osteoclasts exhibited characteristic podosome belts; 10 μM NC9 treatment significantly reduced this ratio to 8%, while 5 ng/ml Exoenzyme C3 brought this back to 67% (Figures 7d and 7e). Taken together, our data demonstrate that TG activity plays an important role in regulation of RhoA and actin dynamics in osteoclast.

TG activity does not affect microtubule architecture or dynamics in osteoclasts Studies have shown that microtubules regulate the expansion of podosome clusters into a belt and that thus disruption of the microtubule network can result in the loss of podosome belt [148,159]. Furthermore, previous studies showed that Exoenzyme C3 increased the acetylated α-tubulin level in osteoclasts [159], and our group has shown that NC9 regulates detyrosinated Glu-tubulin in osteoblasts [286,452]. Examination of microtubule architecture in osteoclast after NC9 treatment for 24 hr, however, showed no major alterations (Supplementary Figure S4b) and western blotting analysis confirmed that NC9 and Exoenzyme C3 had no effect on the acetylation or detyrosination of α-tubulin in osteoclasts (Supplementary Figure S4c).

Expression of TG1, TG2, and FXIII-A by osteoclasts and their localization to podosome belt Next, we were interested in understanding the levels of TG activity during osteoclastogenesis and which TG enzyme(s) are expressed and potentially involved in this process. To measure TG activity and screen TG (Tgm) family member gene expression, we collected cell lysates

64 and mRNA from bone marrow macrophages (day 0), macrophages (day 6), and osteoclasts (day 6). As seen in Figure 8, TG activity was present in bone marrow macrophages and dramatically increased after 6 days of M-CSF treatment, and then decreased with joint treatment with M-CSF and RANKL. RT-PCR analyses of TG family members: Tgm1, Tgm2, Tgm3, Tgm5, Tgm6, Tgm7, and F13a1 (band 4.2 and Tgm4 were not screened because they are exclusively expressed in red blood cells and prostate, respectively [190], showed clear presence of only Tgm1, Tgm2, and F13a1 genes (Supplementary Figure S5). Quantification of their mRNA expression with qPCR analysis revealed that Tgm2 mRNA was expressed at fairly high levels which remained unaltered upon differentiation cues (Figure 9a). F13a1 was significantly upregulated upon M-CSF treatment, but significantly and dramatically reduced with joint M- CSF/RANKL treatment (Figure 9a). Tgm1 was significantly and dramatically upregulated by M-CSF and downregulated when RANKL was added (Figure 9a). The cellular localization of TG1, TG2, and FXIII-A proteins were assessed by immunofluorescence microscopy in osteoclasts. All three enzymes demonstrated a similar distribution pattern: presence in the nuclear region and in the peripheral podosome belt (Figure 9b). Co-staining of TGs with F- actin confirmed that all the three enzymes co-localized to podosomes (Figure 9c). Isotype control staining for TG1, TG2 and FXIII-A antibodies were negative (Supplementary Figure S6). In summary, our work confirms the new role for TGs and TG activity in osteoclastogenesis. TG activity is needed for differentiation, migration and fusion of osteoclasts and regulates actin dynamics in these cells.

3.5 Discussion Previous studies, including ours, have linked TG enzymes and their activity to monocyte, macrophage, and osteoclast function in vitro and in vivo [288-290,295,296,323,356] and our recent work demonstrated the relevance of TG2, FXIII-A in regulation of osteoclastogenesis in mice as well as showed for the first time the presence of TG1 in osteoclasts [288]. In the present study, we have continued the exploration on the role of TGs in osteoclastogenesis and show that TG inhibitors are highly potent blockers of osteoclast differentiation, fusion and migration. These functions for TG activity are aligned with several previous findings on the involvement of TGs in signaling pathways related to osteoclastogenesis, cell fusion, and migration. Inhibition of TGs/TG activity by NC9 blocked macrophage differentiation to osteoclasts. Analysis of osteoclast markers showed that TG inhibitor significantly downregulated Nfatc1,

Trap, Cstk, and Mmp9 in the first two days of treatment with M-CSF and RANKL which induces differentiation from bone marrow macrophages to pre-osteoclasts. NC9 did not have significant effect on these genes after the initial differentiation occurred, that is, fusion stage between day 3 and 6. Pre-osteoclast fusion to multinucleated mature osteoclasts is regulated by master fusogenic factors Dcstamp and Ocstamp [135,137,139,433] which were both significantly downregulated by NC9 in the first two days of treatment with M-CSF and RANKL. As it is well known that NFATc1 is the master regulator of osteoclast differentiation and transcriptionally regulates Trap [120], Ctsk [431], Mmp9 [432], Dcstamp and Ocstamp [121,138], the inhibited expression of Nfatc1 may be responsible for the decreased expression of the others. The difference in the expression levels of the markers in NC9 non-treated cells reflects the different stages of the cells in Exp 1 (end point at day 2) and Exp 2 (end point day 6) as illustrated in Figure 2a. Number of studies showed similar gene expression patterns at early and late differentiation stages [453-455]. The observation that NC9 downregulated Nfatc1 seemingly contradicts to the work of Kim et al., which showed that TG2 knockdown led to elevated Nfatc1/NFATc1, at both mRNA and protein level, and enhanced NFATc1 transcriptional activity in osteoclasts [290]. However, in their study, except for the weakly detected FXIII-A mRNA, TG2 was the only TG family member expressed at a high level in macrophages and pre-osteoclasts, whereas our study showed clear presence of TG1, TG2 and FXIII-A during differentiation and more importantly, the induction of TG1 and FXIII-A with M-CSF. Furthermore, their study did not examine the effect of TG2 knockdown on the expression of other TGs or TG activity in general. Our work shows that TG activity is required for pre-osteoclast migration. TGs, particularly TG2 has been linked to cell migration in several previous publications [238,323,449,450]. Akimov et al. showed cell surface TG2 acted as an integrin-associated adhesion receptor involved in extravasation and migration of monocytes into tissues [323]. Monteagudo et al. showed that ablation of TG2 significantly inhibited migration of primary astrocytes and this migration defect was only rescued with the native protein, not with mutants lacking activity [449] strongly suggesting that TG activity of TG2 is driving migration (versus other non- enzymatic functions of TG2). It was also reported that TG2 can affect cell migration through cross-linking PDGF-BB [450], which interestingly is a highly potent chemotactic factor known to promote monocyte migration [456-458]. However, it is likely that the effects we see in our work arise from TG2 function in podosomes where it localizes. In our work we show that TG inhibition by NC9 increases RhoA levels in cells and inhibits

66 podosome belt formation in osteoclasts. Inhibitor of RhoA-family, Exoenzyme C3, reverses all effects of NC9 strongly suggesting that mechanism of action of TGs in osteoclastogenesis involved Rho GTPases. RhoA, one of the best studied members of Rho GTPases, has been found to control podosome patterning in osteoclast [158]. Osteoclasts on mineralized surfaces form a sealing zone that allows isolated resorption to occur, however, when osteoclasts are seeded on glass surfaces they do not form a sealing zone, but a podosome belt as in our study [147]. The requirement for RhoA activity was shown to be different for the two surface phenotypes (sealing zone versus podosome belt) of osteoclasts, that is, high basal RhoA activity was required for sealing zone formation [150] but it did not trigger sealing zone formation in osteoclasts seeded on glass [157,159]. On the other hand, RhoA inhibition accelerated podosome belt formation in osteoclasts seeded on glass [159], but disrupted the sealing zone and triggered the podosome belt formation in osteoclasts seeded on bone [150], indicating that lowering the RhoA activity in cells is required for podosome belt formation. In addition, a previous study showed that upregulated expression of RhoA suppressed osteoclastogenesis and treatment with the Rho-family inhibitor (Exoenzyme C3) rescued the impaired osteoclastogenesis [459]. In this study, we show that inhibition of TGs increases RhoA levels in cells. Furthermore, we show here that TG1, TG2 and FXIII-A co-localize to the podosomes where RhoA is also found in osteoclasts [147]. Although the precise mechanism by which TGs may regulate RhoA activity (or other Rho-family members) in osteoclasts is likely complex and remains to be explored, previous findings have shown several links of TGs to RhoA. It was shown that TG2 can promote RhoA activation through integrin clustering and inhibition of the Src-p190RhoGAP signaling pathway in fibroblasts [460]. Also, Guilluy et al. have shown that TG2-mediated serotonylation leads to RhoA activation and increased proteasomal degradation in vascular smooth muscle cells [461]. This activation of RhoA due to TG2-mediated serotonylation was also involved in pulmonary artery smooth muscle cell proliferation [462]. We have also demonstrated that NC9 can target FXIII-A in preadipocytes and affect RhoA/ROCK activity [298]. No data is available on TG1 and Rho GTPases. While these findings where TG2 is shown to activate RhoA might seem contradictory to our current study, there are also several studies showing that TG2 can increase the levels of active Rac1 (through inhibiting the Bcr GTPase-activating activity in COS-1 cells [463]) which is known to antagonize RhoA function in osteoclasts [157]. Therefore, TG2 may promote Rac1 activity and inhibit RhoA activity in osteoclasts. However, all above data on TG2 and RhoA can only very carefully be extrapolated to our work where three TGs appear involved in the process and other

Rho-family members might contribute to the effects. More work need to be done to explore which TGs and how they control RhoA or Rho-family activity in macrophages and osteoclasts. This is the focus of our future work. Our previous study demonstrated that the TG enzymes expressed in osteoclasts are TG2, FXIII- A and TG1 and here we also confirm their mRNA expression and cellular localization and show their differential expression patterns during osteoclastogenesis. Our data show that TG2 is expressed at similar levels in bone marrow macrophages (day 0), macrophages (day 6), and osteoclasts (day 6) with no apparent upregulation, whereas FXIII-A is expressed at high levels in bone marrow macrophages (day 0) and macrophages (day 6), but downregulated dramatically in osteoclasts (day 6). TG1 in turn shows a significant increase with M-CSF treatment in macrophages (day 6) and with M-CSF and RANKL treatment in osteoclasts (day 6), but the upregulation by M-CSF alone is higher than the M-CSF/RANKL treatment. These different expression patterns suggest that the three TGs may play distinct roles in osteoclastogenesis. In our study, we found that inhibitors NC9 and MDC were more potent than inhibitors Z006 (inhibits TG2) and T101 (inhibits TG2 and FXIII-A). Collectively, the inhibitor data may suggest the following: 1) Z006 and T101 may target specific TGs at lower concentrations but inhibit all the three TGs at higher concentrations; 2) mechanism of action likely involves protein cross-linking (in addition to potentially involving deamination reaction or non-enzymatic functions of TGs [188,464]) because MDC, competitive inhibitor of protein cross-linking, was able to block osteoclastogenesis. Our future studies will include dissecting out these functions. The discovery of TG1 in macrophages and osteoclasts in our previous work and in this study suggests that it has a role in bone remodeling. TG1, also named as keratinocyte TG, is responsible for the formation of cornified cell envelope that acts as protective skin barrier [190]. Mutations in TGM1 causes lamellar ichthyosis in humans, an autosomal recessive skin disorder [299-301]. Tgm1/TG1 expression has not been previously documented in any bone cells prior to our work [288] and no skeletal phenotype has been reported for Tgm1 deficient mice which die within 4-5 hr after birth [273]. In conclusion, our study describes a role for TGs in osteoclastogenesis. Inhibitors of TGs may provide excellent anti-resorptives that could target all stages of osteoclastogenesis.

3.6 Acknowledgments The authors thank Aisha Mousa for helpful assistance and Dr. Svetlana Komarova (Faculty of Dentistry, McGill University) for helping with osteoclast cultures. This study was supported

68 by grants to MTK from the Canadian Institutes of Health Research (CIHR) (MOP-119403). HS is supported by stipend from the Faculty of Dentistry of McGill University. MTK is members of the Fonds de Recherche-Santé (FRQ-S) Network for Oral and Bone Health Research.

3.7 Figures FIGURE 1

a M-CSF (D5) (-ve ctrl) M-CSF + RANKL (D5) + 5μM NC9 (D5) + 10μM NC9 (D5)

+ 20μM NC9 (D5) + 30μM NC9 (D5) + 40μM NC9 (D5) + 50μM NC9 (D5)

b c d

1.2 **

120 *** 120 1 * ** 100 100 0.8

80 80 steoclast steoclast

0.6 *

60 60 of o of 0.4 *** 40

osteoclast 40

***

*** ***

0.2 20 20 ***

***

*** ***

TRAP activity (OD activity 540) TRAP

***

Number Number

Number of nuclei per nuclei Number of *** 0 0 0 *** 0 10 20 30 40 50 -ve 0 5 10 20 30 40 50 -ve 0 5 10 20 30 40 50 Ctrl Ctrl NC9 concentration (μM) NC9 concentration (μM) NC9 concentration (μM)

FIGURE 1 Inhibition of TG activity with NC9 blocks osteoclastogenesis. (a) The effect of NC9 on osteoclast differentiation. Bone marrow macrophages were treated with M-CSF and RANKL as well as different concentrations of NC9 for 5 days followed by TRAP staining at end point. Magnification bar represents 500 μm. (b) TRAP activity in the culture supernatant is reduced significantly by NC9 in a dose-dependent manner. (c) The number of TRAP+ multinucleated (≥3 nuclei) osteoclasts is first increased followed by a dramatic decrease and complete blockage of osteoclastogenesis. (d) Number of nuclei (visualized by DAPI staining) per osteoclast in the presence of NC9 shows dose-dependent decrease in osteoclast size. Error bars represent SEM. n = 3. *p < 0.05, **p < 0.01, ***p < 0.001.

70 FIGURE 2

a Osteoclastogenesis Differentiation Fusion D0 D2 D5 Exp 1

± NC9 TRAP

staining CSF + + RANKL CSF

- Exp 2 M

± NC9 TRAP staining

b M-CSF + RANKL (Ctrl) + 20μM NC9

Exp Differentiation

M-CSF + RANKL (Ctrl) + 20μM NC9

Exp Fusion

c Exp 1 Exp 2 Exp 2 70 70 2000 NS

60 60 per per 1500 50 50

osteoclast 40 40 osteoclast 1000

30 30 osteoclast osteoclast

*** 20

500 20

Number of nuclei of nuclei Number

Number of of Number Number of of Number 10 10 *** 0 0 0 Ctrl +NC9 Ctrl +NC9 Ctrl +NC9

71 FIGURE 2 TG inhibition blocks pre-osteoclast differentiation and fusion. (a) Schematic diagram of the two phases of osteoclastogenesis, differentiation and fusion, and the experimental design for the NC9 treatments to dissect out these two functions. Exp 1 (Bone marrow macrophages treated with M-CSF and RANKL with 20 μM NC9 for the first 2 days) examines the effect of NC9 on the differentiation of bone marrow macrophages into pre-osteoclasts. Exp 2 (Pre-osteoclasts treated with M-CSF and RANKL with 20 μM NC9 for the last 3 days) examines the effect of NC9 on the fusion of pre-osteoclasts. (b) TRAP staining of osteoclasts formed with NC9 treatment in Exp 1 and Exp 2. NC9 blocks both differentiation and fusion of pre-osteoclasts. Magnification bar represents 500 μm. (c) Quantification of TRAP+ osteoclasts in Exp 1 and Exp 2 and number of nuclei per osteoclast in Exp 2 showing that NC9 significantly inhibits differentiation and fusion of pre-osteoclasts. In Exp 1, the number of pre- osteoclasts is significantly decreased. In Exp 2, osteoclast number remains the same and only size is significantly affected. Error bars represent SEM. n = 3. ***p < 0.001. NS, not significant.

72 FIGURE 3

a Nfatc1 b Trap 1.2 1.8 NS 1.6 1 *** +NC9 1.4 0.8 1.2 *** 1 0.6 NS 0.8 0.4 +NC9 0.6 +NC9

+NC9 0.4 Relative Relative expression Relative Relative expression 0.2 0.2 0 0 DMSOExpNC91 D0- DMSOExpNC92 D2- DMSOExp 1NC9 DMSOExp 2NC9 D0-D2 D2 D2-D6 D6 D0-D2 D0-D2 D2-D6 D2-D6

c Ctsk d Mmp9 1.2 2.5 NS *** 1 2 +NC9 0.8 1.5 0.6 NS *** 1 0.4 +NC9 +NC9 Relative Relative expression 0.5 +NC9 0.2 expression Relative

0 0 DMSOExp 1NC9 DMSOExpNC92 DMSOExp NC91 DMSOExp NC92 D0-D2 D0-D2 D2-D6 D2-D6 D0-D2 D0-D2 D2-D6 D2-D6 e f Dcstamp Ocstamp 1.2 1.2 ** *** 1 * 1

0.8 +NC9 0.8 0.6 +NC9 0.6 NS 0.4 0.4 +NC9 +NC9

Relative Relative expression 0.2 Relative Relative expression 0.2

0 0 DMSOExp 1NC9 DMSOExpNC92 DMSOExp 1NC9 DMSOExp 2NC9 D0-D2 D0-D2 D2-D6 D2-D6 D0-D2 D0-D2 D2-D6 D2-D6

FIGURE 3 TG inhibition blocks expression of differentiation and fusion markers during osteoclastogenesis. qPCR analysis of osteoclast differentiation and fusion markers using the same treatments in Exp 1 (D0-2) and Exp 2 (D2-6) depicted in Figure 2a. (a) Nfatc1 (NFATc1), (b) Trap (TRAP), (c) Ctsk (CTSK), and (d) Mmp9 (MMP-9) expression were significantly decreased during early differentiation of bone marrow macrophages to pre- osteoclasts. Analysis of expression of fusion markers showed (e) significant Dcstamp (DC- STAMP) decrease in both differentiation and fusion stages, and (f) significant Ocstamp (OC-STAMP) decrease only in early differentiation phase. All the gene expressions were normalized to Gapdh. Error bars represent SEM. n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. NS, not significant.

73 FIGURE 4 a M-CSF + RANKL + 20μM MDC + 40μM MDC 0.8 MDC

) 0.7

540 0.6 *** 0.5

0.4 *** + 60μM MDC + 80μM MDC + 100μM MDC 0.3

0.2

***

*** *** TRAP activity (OD activity TRAP 0.1 0 0 20 40 60 80 100 MDC concentration (μM)

b M-CSF + RANKL + 40μM Z006 + 80μM Z006 1 Z006

0.8 ***

0.6 ***

0.4 + 120μM Z006 + 160μM Z006 + 200μM Z006 ***

0.2 ***

*** TRAP activity (OD activity 540) TRAP

0 0 40 80 120 160 200 Z006 concentration (μM) c M-CSF + RANKL + 200μM T101 + 400μM T101 0.8 T101 ) 0.7 *

540 0.6 0.5 *** 0.4 + 600μM T101 + 800μM T101 + 1mM T101

0.3 *** *** 0.2 ***

TRAP activity (OD activity TRAP 0.1 0 0 200 400 600 800 1000 T101 concentration (μM)

FIGURE 4 TG inhibitors MDC, Z006 and T101 block osteoclastogenesis. TG inhibitors monodansyl cadaverine (MDC) (competitive inhibitor for protein crosslinking), Z006 (TG2 inhibitor) and T101 (TG2 and FXIII-A inhibitor) were tested on osteoclast differentiation. Bone marrow macrophages were treated with M-CSF and RANKL with different concentrations of inhibitors for 5 days followed by TRAP staining and colorimetric TRAP quantification in the culture supernatant. MDC (a), Z006 (b), and T101 (c) were able to inhibit osteoclast fusion at lower concentrations and block osteoclast differentiation completely at higher concentrations. Magnification bar represents 500 μm. Error bars represent SEM. n = 3. *p < 0.05, ***p < 0.001. 74 FIGURE 5

a D0 M-CSF + RANKL (D5)

+ 5μM NC9 (D5) + 10μM NC9 (D5)

+ 20μM NC9 (D5) + 30μM NC9 (D5)

+ 40μM NC9 (D5) + 50μM NC9 (D5)

b 40%

) 30%

rea

a surface surface 20% ** ** **

(percent (percent 10% Resorbed Resorbed

0% D0 0 5 10 20 30 40 50 D5 / NC9 concentration (μM)

FIGURE 5 TG activity regulates osteoclast resorption. (a) Bone marrow macrophages were treated with M-CSF and RANKL for 5 days, followed by addition of different concentrations of NC9 together with M-CSF and RANKL for another 5 days. The resorption pits (white) were visualized with von Kossa staining and photographed. Magnification bar represents 1 mm. (b) Percentages of surface resorbed by osteoclasts at various NC9 concentrations were quantified with NIH ImageJ software. Results showed osteoclast resorption was visibly and significantly and dose-dependently reduced by NC9. Error bars represent SEM. n = 3. **p < 0.01. 75 FIGURE 6

a 0h 12h b 1400

1200 cells

CSF + + RANKL CSF 1000

- M 800 0h 12h 600

400

NC9

Number of migrated migrated Number of M M

μ ***

200

CSF+RANKL

+ + M 0 Ctrl +NC9

FIGURE 6 TG inhibitor NC9 inhibits pre-osteoclast migration. (a) Pre-osteoclast migration was investigated using a cell exclusion zone assay. Pre-osteoclasts were formed in a well with a cell seeding stopper to generate a central exclusion zone after which cells were treated with M-CSF and RANKL in the presence or absence of 20 μM NC9 for 12 hr. Plates were photographed at 0 and 12 hr time points. Magnification bar represents 1100 μm. (b) The number of cells migrated to the exclusion zone during the 12 hr was calculated. Results showed that NC9 inhibited the pre-osteoclast migration in a significant manner. Error bars represent SEM. n = 3. ***p < 0.001.

76 FIGURE 7

a b Total RhoA c RhoA activity 1.5 100 NS Ctrl +NC9 * 80 Total RhoA 1 60 Actin 0.5 40

Relative Relative expression Relative Relative expression 0 0 Ctrl NC9 Ctrl NC9 d + 10μM NC9

M-CSF + RANKL (Ctrl) + 10μM NC9 + 5 ng/ml Exoenzyme C3

TRAP

DAPI Actin Actin +

NS NS NS e 100% 140 160 ** * *** * 80% *** *** 120 140 120 100 100 (%)belt 60% 80 80 40% 60 60 40 osteoclast 40 Osteoclasts Osteoclasts with 20%

20 20 podosome Number osteoclast Number of 0 per nuclei of Number 0 0% Ctrl NC9 NC9 Ctrl NC9 NC9 Ctrl NC9 NC9 + C3 + C3 + C3

77 FIGURE 7 TG inhibitor NC9 increases RhoA levels, decreases podosome belt formation and the inhibitory effect is reversed with a Rho-family inhibitor. (a) Bone marrow macrophages were treated with M-CSF and RANKL in the presence or absence of 20 μM NC9 for 2 days. Western blotting analysis showed that NC9 increased total RhoA levels. (b) Quantification of protein bands showed NC9 significantly increased total RhoA level. The band densities were normalized to the actin loading control. Error bars represent SEM. n = 3. *p < 0.05. (c) The GTP-bound RhoA was pulled down from whole-cell extracts using Rhotekin RBD beads and analyzed by western blotting (not shown). Quantification of protein bands showed elevated RhoA activity, however, this was not statistically significant. Error bars represent SEM. n = 3, NS, not significant. (d) Bone marrow macrophages were treated with M-CSF and RANKL in the presence or absence of 10 μM NC9 and 5 ng/ml Rho-family inhibitor Exoenzyme C3 for 5 days, followed by TRAP staining (Magnification bar represents 500 μm), F-actin and DAPI staining (Magnification bar represents 400 μm). (e) Quantification of number of osteoclasts, size of osteoclasts (number of nuclei per osteoclast) and percentage of osteoclasts with podosome belts. Adding Rho-family inhibitor together with NC9 completely normalized the size of osteoclasts, osteoclastogenesis and podosome belt formation. Error bars represent SEM. n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. NS, not significant.

78 FIGURE 8

D0 D6 D6 D0 D6 D6 M-CSF Cell lysate - + + - + + - - + - - + RANKL Anti- biotin Anti-

assay pFN TG activity activity TG Actin Cell lysate - - - + + + + + + pFN - + + - - - + + + bPA + - + - - - + + +

FIGURE 8 In vitro TG activity in bone marrow macrophages and osteoclasts. TG activity in cell lysates was assessed via primary amine, 5-(biotinamido)pentylamine (bPA) incorporation to a TG substrate protein plasma fibronectin (pFN) in a 2-hr incubation. This was followed by detection of biotin in pFN by western blotting. Cell lysates were prepared from bone marrow macrophages (Day 0), macrophages (Day 6, M-CSF treatment) and osteoclasts (Day 6, M-CSF + RANKL treatment). Assay included negative controls where bPA or pFN or cell lysates were excluded (first six lanes). TG activity is seen in last three lanes. It is detected in bone marrow macrophages and dramatically increased after 6 days of M-CSF treatment and then decreased with joint treatment with M-CSF and RANKL.

79 FIGURE 9

a Tgm2 (TG2) F13a1 (FXIII-A) Tgm1 (TG1) *** * 1.8 NS 1.6 *** 8 ** * 1.6 1.4 ** 7 1.4 1.2 6

1.2 mRNA mRNA mRNA mRNA 1 1 mRNA 5 0.8 0.8 4 0.6

expression 3

expression 0.6

expression Relative Relative

Relative Relative 0.4 0.4 Relative 2 0.2 0.2 1 0 0 0 D0 D6 D6 D0 D6 D6 D0 D6 D6 M-CSF - + + M-CSF - + + M-CSF - + + RANKL - - + RANKL - - + RANKL - - + b c TG2 TG2 Actin Merge

Actin Merge

FXIII-A FXIII-A

CSF+RANKL (D5) CSF+RANKL

M M

TG1 TG1 Actin Merge

80 FIGURE 9 TG enzyme expression in bone marrow macrophages and osteoclasts and their localization in osteoclasts. (a) qPCR analysis of TG2, FXIII-A and TG1 isoforms in bone marrow macrophages (Day 0), macrophages (Day 6, M-CSF treatment) and osteoclasts (Day 6, M-CSF + RANKL treatment). Tgm2 mRNA was expressed at fairly high levels which remained unaltered upon differentiation cues. F13a1 was significantly upregulated upon M-CSF treatment, but significantly and dramatically reduced with M-CSF and RANKL treatment. Tgm1 was significantly upregulated by both of the differentiation cues, but the upregulation induced by M-CSF alone was significantly higher than the joint treatment of M-CSF and RANKL. All the gene expression were normalized to Gapdh. Error bars represent SEM. n = 3. *p < 0.05, **p < 0.01, ***p < 0.001. NS, not significant. (b) Bone marrow macrophages were treated with M-CSF and RANKL for 5 days. TG1, TG2, and FXIII-A in osteoclasts were visualized with their primary antibodies and Alexa Fluor® 568 (red). Nuclei were stained by DAPI (blue). TG2, FXIII-A, and TG1 (red) are found in the edge/periphery of osteoclasts. Magnification bar represents 200 μm. (c) F-actin was visualized with Alexa Fluor® 488 (green). TG2, FXIII-A, and TG1 (red) co-localize with actin (green) (merge in yellow) and are all found in podosomes (arrows). Magnification bar represents 20 μm.

81 SUPPLEMENTAL FIGURE S1

a NC9 b MDC 2.5 ** 1.6 * * NS 1.4 * ** 2 1.2 * NS *

NS 1 NS

OD560)

OD560) ( ( 1.5 0.8 1 NS 0.6 0.4

0.5 Cell viability Cell viability Cell viability Cell viability **** 0.2 0 0 0 5 10 20 30 40 50 60 80 100 0 20 40 60 80 100 NC9 concentration (μM) MDC concentration (μM)

c Z006 d T101 2.5 2.5 ** ** *** 2.0 ** ** ** 2 *

NS ** OD560)

OD560) NS

( 1.5 ( 1.5

1.0 1

0.5 0.5

Cell viability Cell viability Cell viability Cell viability

0.0 0 0 40 80 120 160 200 0 200 400 600 800 1000 Z006 concentration (μM) T101 concentration (μM)

FIGURE S1 The effect of TG inhibitors, NC9, MDC, Z006 and T101 on cell viability in osteoclasts. Bone marrow macrohages were treated with M-CSF and RANKL in the presence of different concentrations of the inhibitors for 5 days. The cell viability was measured by MTT assay which measures number of cells per well. (a) Cells tolerated NC9 well up to 50-60 µM level. (b) Cells tolerated MDC up to the maximum 100 mM tested. (c) Cells tolerated Z006 up to maximum 200 mM tested. (d) Cells tolerated T101 up to maximum 1 mM tested. The apparent increase in cell numbers is likely due to proliferation of macrophages which was permitted once osteoclastogenesis was inhibited. Error bars represent SEM. n = 3. *p< 0.05, **p< 0.01, ***p< 0.001. NS: not significant.

82 SUPPLEMENTAL FIGURE S2

a D0 b M-CSF (D5)

d c M-CSF + RANKL (D2) M-CSF + RANKL (D3) e M-CSF + RANKL (D5)

FIGURE S2 TRAP staining images of different stages of osteoclasts and their precursors. (a) M-CSF-dependent bone marrow-derived macrophages were used as osteoclast precursors at day 0. (b) Cells were treated with M-CSF for 5 days to serve as negative control. (c) Cells were treated with M-CSF and RANKL for 2 days (D2) to form TRAP+ mononuclear pre-osteoclasts. (d) Cells were treated with M-CSF and RANKL for 3 days (D3) to form TRAP+ multinucleated osteoclasts. (e) Cells were treated with M-CSF and RANKL for 5 days (D5) to form mature osteoclasts. Cells were fixed and stained for TRAP with the TRAP Staining Kit. Magnification bar represents 500 μm.

83 SUPPLEMENTAL FIGURE S3

a D0 b M-CSF + RANKL (D5) + 40μM NC9 (D5)

FIGURE S3 TRAP staining images of osteoclasts during osteoclastic resorption assay. (a) Bone marrow macrophages were plated into the Corning® Osteo Assay Surface 96- well plates and osteoclastogenesis were induced by M-CSF and RANKL. After 5 days, mature osteoclasts were fixed and stained for TRAP with the TRAP Staining Kit. (b) Mature osteoclasts were cultured for further 5 days (D5) with M-CSF and RANKL in the presence or absence of NC9. Cells were fixed and stained for TRAP. Magnification bar represents 500 μm.

84 SUPPLEMENTAL FIGURE S4 a M-CSF + RANKL

M-CSF + RANKL + 10 μM NC9

DAPI Actin Actin +

b M-CSF+ RANKL + 20μM NC9

DAPI

+ +

Tubulin

- 

c Ctrl NC9 NC9 + C3 Ac-tub

Glu-tub

-tubulin

FIGURE S4 The effect of NC9 on podosome belt formation and microtubule architecture and dynamics in osteoclasts. (a) Bone marrow macrophages were treated with M-CSF and RANKL in the presence or absence of 10 μM NC9 for 5 days. Cells were fixed and stained for F-actin and DAPI. Upon NC9 treatment, cells lost ability to form podosome belts and only actin clusters were found. Magnification bar represents 400 μm (left). Magnification bar represents 50 μm (right). (b) Bone marrow macrophages were treated with M-CSF and RANKL for 4 days after which cells were treated with M-CSF and RANKL in the presence or absence of 20μM NC9 for 24 hours. Osteoclasts were fixed and stained with anti--tubulin antibody and then Alexa Fluor® conjugated secondary antibody. Nuclei were stained by 4′,6-diamidino-2- phenylindole (DAPI). No apparent change to microtubule architecture was observed upon NC9 treatment. Magnification bar represents 200 μm. (c) Bone marrow macrophages were treated with M-CSF and RANKL in the presence or absence of 10μM NC9 and 5 ng/ml Exoenzyme C3 for 6 days. Total cellular proteins were extracted and analyzed with Western blotting using antibodies against acetylated tubulin (Ac-tub), detyrosinated Glu-tubulin (Glu-tub) and - tubulin. No apparent change to the acetylation and detyrosination of -tubulin was observed upon NC9 treatment or NC9 + C3 treatment. Ctrl represents vehicle control.

85 SUPPLEMENTAL FIGURE S5

Tgm1 Tgm2 Tgm3 Tgm5 Tgm6 Tgm7 F13a1 Gapdh

D6 / M-CSF

D6 / M-CSF + RANKL

FIGURE S5 RT-PCR analysis of TG family members bone marrow macrophages (D0), macrophages (D6) and osteoclasts (Day 6). Only Tgm1, Tgm2 and F13a1 were expressed. n = 2.

86 SUPPLEMENTAL FIGURE S6

a Mouse IgG2a DAPI Merge

b Mouse IgG1 DAPI Merge

c Rabbit IgG DAPI Merge

FIGURE S6 The isotype controls for antibodies against TG1, TG2 and FXIII-A used in FIGURE 9. Bone marrow macrophages were treated with M-CSF and RANKL for 5 days to form osteoclasts. Cells were fixed and stained with (a) Mouse IgG2a (isotype control for TG1), or (b) Mouse IgG1 (isotype control for TG2), or (c) Rabbit IgG (isotype control for FXIII-A), and then Alexa Fluor® conjugated secondary antibody. Nuclei were stained by 4′,6-diamidino- 2-phenylindole (DAPI). Magnification bar represents 400 μm.

87 Chapter 4 - Assessment of expression and specific activities of transglutaminases TG1, TG2 and FXIII-A during osteoclastogenesis

Recent studies have shown that FXIII-A deficiency results in reduced osteoclastogenesis in vivo and in vitro, and TG2 deficiency gives rise to increased osteoclastogenesis in vivo and in vitro. We have shown that three TGs, TG1, TG2, and FXIII-A are present in osteoclasts and their precursors and Tgm2-/-;F13a1-/- monocytes have increased potential to form osteoclasts in vivo and in vitro. Furthermore, we also have shown using an array of TG inhibitors that TG activity is required for osteoclastogenesis. However, the total and individual activities of the three TGs during osteoclastogenesis and their regulation by M-CSF and RANKL have not been examined yet. Thus, in this study, I examined the temporal expression and activities of TGs during osteoclastogenesis on a day-by-day basis. Total TG activity was assessed using 5- (Biotinamido)pentylamine (bPA) and specific activities of TG1, TG2, and FXIII-A were analyzed using biotinylated “Hitomi peptides”, bK5, bT26 and bF11. I report that TGs activities are highest in the differentiation and early fusion phases and then decrease dramatically. TGs activities were upregulated by M-CSF and downregulated by addition of RANKL. FXIII-A was dramatically downregulated by RANKL, suggesting its involvement in M-CSF-mediated precursor commitment phase. TG1 and TG2 proteins were present throughout osteoclastogenesis, suggesting that they may have functions in both differentiation and fusion. In summary, the three TGs likely exert distinct functions at different stages of osteoclastogenesis. Our work also demonstrates that the “Hitomi peptides” are highly specific tools for detection of distinct TGs in a system where multiple TGs are present. It is the first report describing the total and separate activities of TGs and their regulation by M-CSF and

RANKL during osteoclastogenesis.

The study presented in this chapter has been submitted to a special issue on “Transglutaminases in Translation - Novel Tools and Methods Impacting on Diagnostics and Therapeutics” in a peer-reviewed journal Analytical Biochemistry.

Assessment of expression and specific activities of transglutaminases TG1, TG2 and FXIII-A during osteoclastogenesis

Sun H1 and Kaartinen MT1,2,3 1 Division of Biomedical Sciences, Faculty of Dentistry, McGill University, Montreal, QC, Canada 2 Division of Experimental Medicine, Department of Medicine, Faculty of Medicine, McGill University, Montreal, QC, Canada 3Department of Anatomy and Cell Biology, Faculty of Medicine, McGill University, Montreal, QC, Canada

4.1 Abstract

Osteoclasts are large multinucleated bone-resorbing cells derived from monocyte/macrophage lineage. Macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor-κB ligand (RANKL) drive the multi-stage osteoclastogenesis. Transglutaminases (TGs) are Ca2+- and thiol-dependent acyl transferases and protein crosslinking enzymes. TG enzyme family contains eight catalytically active enzymes TG1-7 and Factor XIII-A (FXIII-A). Recent studies have shown that TG1, TG2, and FXIII-A are present in osteoclasts and that TG2 and FXIII-A regulate osteoclastogenesis. In this study, we examined gene and protein expression and specific activities of TG1, TG2 and FXIII-A during osteoclastogenesis using “Hitomi peptides” in a day-by-day manner. We report that TGs activities are highest in the differentiation and early fusion phases and then decrease dramatically. TGs activities were upregulated by M-CSF and downregulated by addition of RANKL. FXIII-A was dramatically downregulated by RANKL, suggesting its involvement in M-CSF-mediated precursor commitment phase. TG1 and TG2 proteins were present throughout osteoclastogenesis, suggesting that they may have functions in both differentiation and fusion. In summary, the three TGs likely exert distinct functions at different stages of osteoclastogenesis. Our work also demonstrates that the “Hitomi peptides” are highly specific tools for detection of distinct TGs in a system where multiple TGs are present.

4.2 Introduction Osteoclasts are large multinucleated bone-resorbing cells that derive from monocyte/macrophage lineage [15]. Increased osteoclast activity is responsible for bone destruction in diseases such as osteoporosis, periodontitis, and rheumatoid arthritis [416]. Osteoclastogenesis is a complex process that includes several stages, such as precursor commitment, differentiation, and multinucleation, which are driven by two cytokines, macrophage-colony stimulating factor (M-CSF) and receptor activator of nuclear factor-κB ligand (RANKL) [16,465]. M-CSF promotes proliferation and survival of macrophages which act as osteoclast precursors [430]. M-CSF is also necessary for osteoclast differentiation and primes the macrophages to respond to RANKL by inducing expression of its receptor RANK [56]. RANKL, the key osteoclastogenic cytokine, activates NF-кB and MAPKs pathways which subsequently induce the expression of nuclear factor of activated T cell c1 (NFATc1), which is a master regulator of osteoclastogenesis [76]. NFATc1 in turn drives expression of various osteoclast-specific genes such as tartrate resistant acid phosphatase (TRAP) and cathepsin K (CTSK). In addition, NFATc1 also triggers expression of various molecules involved in the fusion of osteoclasts, such as dendritic cell-specific transmembrane protein (DC-STAMP) and d2 isoform of vacuolar ATPase V0 domain (Atp6v0d2) [76]. At very early stages of osteoclastogenesis, F-actin-rich adhesive structures called podosomes are formed. They participate in osteoclast adhesion, spreading, migration, and bone degradation [435]. Transglutaminases (TGs; EC 2.3.2.13) are a family of Ca2+ and thiol-dependent acyl transferases which catalyze covalent cross-linking, i.e., isopeptide formation between the γ- carboxamide group of a protein- or peptide-bound glutamine residue (also known as the “Q” substrate) and an ε-amine group of a protein-bound lysine residue or a primary amine (also known as the “K” substrate) [196]. TG family consists of eight catalytically active enzymes, TG1-7 and Factor XIII-A (FXIII-A) that share 100% homology of their active site residues, yet having high specificity for their substrates [188]. In the presence of Ca2+, TG2 adopts an open conformation, making the catalytic core accessible to the substrates [192]. Among the TGs,

TG1 and FXIII-A have an additional NH2-terminal pro-peptide whose removal promotes their catalytic activation [188]. TG activity requires free thiol group of the critical cysteine residue

90 of the active site, making TGs susceptible to reactive oxygen species-induced or nitric oxide (NO)-induced inactivation [202,206]. Therefore, TG activity can be regulated by Ca2+ levels, proteolysis and by the redox status of the cells. The presence of TGs in bone and bone cells has been well documented via both in vitro and in vivo studies [238,256,280]. While the original work has mostly focused on osteoblasts [281,284-286,437-441], latest work by us and others have revealed their regulatory roles in osteoclastogenesis [288-291]. The work of Raghu et al. showed that FXIII-A deficiency results in reduced osteoclastogenesis in vivo and in vitro [289]. Our recent work demonstrated that double ablation of TG2 and FXIII-A gives rise to increased osteoclastogenesis in vivo and in vitro, which suggests that these two enzymes synergistically and negatively regulate osteoclastogenesis [288]. However, we also demonstrated that wild type and TG2/FXIII-A double null monocytes express TG1, and that a TG inhibitor NC9 was able to block the osteoclast differentiation in both wild type and double null monocytes where increased differentiation was observed. This suggests that TG1 may be an enzyme that promotes osteoclastogenesis [288]. To add another level of complexity, a recent study from Kim et al. showed that TG2 deficiency results in increased osteoclastogenesis in vivo and in vitro [290]. In addition, we showed recently that four different TG inhibitors can block osteoclast differentiation and that TG activity was involved in regulation of multiple steps in osteoclastogenesis via affecting actin dynamics [291]. Based on this robust in vivo and in vitro data, TGs may be interesting therapeutic targets to increase bone mass via limiting aberrant osteoclast activity. As outlined above, the complicating factor in using TGs as targets in osteoclasts is that three TGs, FXIII-A, TG2, and TG1 are present with potential distinct roles in osteoclastogenesis. To be able to target the correct TGs with high specificity, it is important to understand with precision, how and when they function. To date, nothing is known about their expression and activity patterns during osteoclastogenesis. Therefore, the aim of this study was to investigate the expression of FXIII- A, TG2, and TG1 and their specific enzymatic activities using “Hitomi peptides” in a day-by- day manner during the osteoclastogenesis [211-213]. Our study reveals that the three TGs likely exert distinct functions at different stages of osteoclastogenesis, i.e., FXIII-A is likely involved

91 in M-CSF-mediated precursor commitment phase, and both TG1 and TG2 likely play an important role from day 1 to day 3, which are the differentiation and early fusion stages. Activity assays together with gene and protein expression data demonstrate that “Hitomi peptides” can distinctly detect activities of individual TGs in a system where multiple enzymes are present.

4.3 Materials and methods

Reagents and antibodies MEM Alpha (αMEM) (12561-056), penicillin-streptomycin, L-glutamine, and sodium pyruvate were from Gibco (Burlington, ON, Canada). Fetal bovine serum was from Hyclone (Waltham, MA). Human M-CSF and human sRANK Ligand were from PeproTech (Rocky Hill, NJ). 5-(biotinamido)pentylamine (bPA) was from Thermo Fisher Scientific (Rockford, IL).

Biotinylated “Hitomi peptides”, bF11, bT26, bK5, and their control peptides, bF11QN, bT26QN, bK5QN, were synthesized by Biomatik Corporation (Cambridge, ON, CA). NC9 was synthesized by Gene Tech Inc (Indianapolis, IN) [286]. All other reagents unless otherwise specified were purchased from Sigma-Aldrich (Oakville, ON, Canada) or Fisher Scientific (Hampton, NH). Mouse monoclonal TG2 antibody (Ab-1) was from Fisher Scientific. Rabbit polyclonal TG1 antibody (ab103814) and rabbit polyclonal FXIII-A antibody (ab97636) were from Abcam (Cambridge, MA). Rabbit polyclonal actin antibody, mouse monoclonal biotin antibody (BN- 34) were from Sigma-Aldrich. Alexa Fluor® 488-phalloidin and DAPI (4’, 6-diamidino-2- phenylindole) were from Thermo Fisher Scientific. Secondary antibody HRP-linked anti- mouse IgG was from GE Healthcare (Mississauga, ON, Canada). HRP-linked anti-rabbit IgG was from Cell Signaling (Whitby, ON, Canada).

Osteoclast differentiation from mouse bone marrow cells Studies were compliant with McGill University guidelines established by the Canadian Council on Animal Care. Complete medium for this study is composed of αMEM, 10% FBS, 1% penicillin-streptomycin, 1% L-glutamine solution, and 1% sodium pyruvate. Murine bone marrow cells were collected from 6-10 weeks old C57BL/6 mice (from Jackson Laboratories) 92 as described previously [291]. Briefly, bone marrow cells were plated at a density of 15  106 cells per 75 cm2 tissue culture flask and cultured overnight in and 25 ng/ml M-CSF. After 24 hr, non-adherent cells were collected and plated at 5  104 cells/cm2 and incubated with 50 ng/ml M-CSF. After 48 hr, the non-adherent cells were discarded, and the adherent cells were considered to be M-CSF-dependent bone marrow-derived macrophages (BMMs) and used as osteoclast precursors. Osteoclast differentiation was then induced by 50 ng/ml M-CSF and 50 ng/ml RANKL for 6 days with two medium changes on day 2 and day 4. The BMMs were also treated with only 50 ng/ml M-CSF for 6 days to serve as a negative control since only joint treatment with M-CSF and RANKL results in osteoclastogenenesis.

TRAP staining BMMs were plated in 24-well plates and treated with M-CSF and RANKL in the presence or absence of 20 µM NC9. DMSO is used as drug vehicle. At day 6, cells were fixed with 3.7% formaldehyde and stained for TRAP with the TRAP Staining Kit (Cosmo Bio, Carlsbad, CA). Images were taken under 10  objective by using the Leica DMi8 inverted fluorescence microscope and Leica Application Suite X software (version 3) (Leica, Concord, ON, Canada).

TRAP activity assay BMMs were plated on a 24-well plate and cultured as described above. At the end points, TRAP activity in the culture supernatant was measured with the TRAP Staining Kit (Cosmo Bio, Carlsbad, CA) as described previously [291]. Briefly, reaction buffer was prepared according to the manufacturer's instructions and was added to the culture supernatants in a 96-well plate. The plate was then incubated for 3 h in dark at 37°C and then read in a microplate reader at 540 nm (TECAN infinite F200 PRO, Männedorf, Switzerland). The experiments were independently repeated three times, each time in triplicate.

Immunofluorescence microscopy BMMs were plated on a 24-well plate and cultured as described above. At the end points, cells were fixed with 3.7% formaldehyde and blocked with 2% bovine serum albumin (BSA). F-

93 actin was labeled with Alexa Fluor® 488-phalloidin and nuclei were stained by DAPI. Images were taken under 20  objective by using the Leica DMi8 inverted fluorescence microscope and Leica Application Suite X software.

Protein extraction and Western blotting Cell lysates were prepared with ice-cold cell lysis buffer containing 50 mM Tris (pH 7.5), 10 mM MgCl2, 0.5 M NaCl, and 2% Igepal, 1% protease inhibitor cocktail, and 1% phosphatase inhibitor cocktail and incubated on ice for 30 min. Cells were scraped and the lysates were further disrupted by passage through a 25G syringe needle for 30 times on ice and centrifuged at 14,000 g for 15 min at 4 °C. Protein concentrations were measured using the BCA Protein Assay Kit (Thermo Fisher Scientific). For Western blotting, 40 μg proteins (for TG1) or 20 μg proteins (for TG2 and FXIII-A) were separated on 4-20% Mini-PROTEAN TGX precast protein gels in denaturing conditions and transferred to PVDF membrane (Bio-Rad, Mississauga, ON, Canada) followed by blocking with 5% nonfat milk in Tris buffered saline with 0.1% Tween 20 (TBST), and individual proteins were detected with specific antibodies as well as corresponding secondary antibodies conjugated with horseradish peroxidase. Chemiluminescent signal was generated by the ECL Plus kit (GE Healthcare), and then was detected by ChemiDocTM Touch Imaging System (Bio-Rad). The experiments were independently repeated twice.

In vitro TG activity assay In vitro TG activity assay was performed as described previously [291]. Briefly, 20 μg protein was incubated for 2 h at 37 °C with either bPA or biotinylated “Hitomi peptides”, bF11, bT26, bK5, and their control peptides, bF11QN, bT26QN, bK5QN, in a reaction buffer containing 1 mM dithiothreitol (DTT), 3 mM CaCl2, 10 mM Tris-HCl (pH 8.0). bF11, bT26, bK5, bF11QN, bT26QN, bK5QN were dissolved in dimethylsulfoxide (DMSO), and bPA in H2O. The concentrations for all the reagents are listed in Table I. After incubation, samples were analyzed by Western blotting as above to visualize covalent incorporation of bPA or biotinylated “Hitomi peptides” to their substrates which reflects the TG activity present in the samples. The

94 experiments were independently repeated twice.

RNA extraction and qPCR Total RNA from cell cultures was extracted as described previously [291]. Briefly, RNA was extracted using the RNeasy Mini Kit (Qiagen, Venlo, Netherlands), which includes an “on column DNase digestion” with the RNase-Free DNase Set (Qiagen). The concentrations of RNA were determined by the TECAN infinite F200 PRO with the NanoQuant Plate. For qPCR, cDNA was synthesized from RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). qPCR was performed on the StepOnePlus Real-Time PCR System (Applied Biosystems). TaqMan® Fast Advanced Master Mix and primers were purchased from Applied Biosystems. Expression levels of Tgm1 (Mm00498375_m1), Tgm2 (Mm00436987_m1), F13a1 (Mm00472334_m1), Nfatc1 (Mm00479445_m1), Acp5 (Mm00475698_m1), Ctsk (Mm00484039_m1), Dcstamp (Mm04209234_m1), Atp6v0d2 (Mm01222963_m1), Adgre1 (Mm00802529_m1), Arg1 (Mm00475988_m1), Nos2 (Mm00440502_m1) were normalized to Gapdh (Mm99999915_g1). The experiments were independently repeated three times, each time in triplicate.

Statistical analysis Data were analyzed with GraphPad Prism software (version 8.0.1). Results are presented as averages  SEM (standard error of the mean) of three independent experiments done in triplicates. Statistical significance was analyzed with one-way ANOVA, followed by Tukey post test. p values  0.05 were considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, NS: not significant).

4.4 Results Day-by-day examination of macrophage and osteoclast markers

The aim of this study was to examine the expression and activities of FXIII-A, TG2, and TG1 during 6-day osteoclastogenesis to shed light on their roles in osteoclast differentiation. A day- by-day examination approach was used because osteoclast formation from bone marrow- derived macrophages (BMMs) involves two main stages where initial differentiation is 95 followed by migration and fusion [16,465]. BMMs were treated with either M-CSF alone or M-CSF and RANKL for 6 days, former treatment maintains macrophage phenotype and latter induces osteoclastogenesis. Macrophages respond to various stimuli by adopting a specific activation phenotype, i.e., classically activated (M1) and alternatively activated (M2) macrophages [30]. M-CSF is linked with M2 macrophage phenotype, however, it is not a potent stimulus for definitive polarization markers compared with prototypic polarizing stimuli such as IL-4 and IL-13 [466]. The pan-macrophage marker Adgre1 (F4/80) [467] was upregulated by the treatment and peaked on day 6. M2 macrophage marker Arg1 (ARG1) [467] was downregulated by the treatment and was the lowest at day 2, 4 and 6, whereas M1 macrophage marker Nos2 (iNOS) [467] was upregulated by the treatment but the increase was only significant at day 6 (Figure 1). This suggests that M-CSF-treated BMMs likely adopt the M1 macrophage phenotype. Since it has been shown that RANKL could trigger M1-like macrophages with properties that were different from those of LPS and IFN-γ-induced macrophages [468], we also examined the effect of joint M-CSF and RANKL treatment on the expression of these macrophage markers during osteoclastogenesis. Figure S1 showed that M- CSF and RANKL induced the similar expression pattern of macrophage markers in BMMs to M-CSF. The pan-macrophage marker Adgre1 (F4/80) was upregulated by the treatment and peaked on day 4. M2 macrophage marker Arg1 (ARG1) was downregulated by the treatment with the lowest expression on day 2 and 4, whereas M1 macrophage marker Nos2 (iNOS) was upregulated dramatically by the treatment and peaked on day 6. This may have relevance to regulation of TG activity in osteoclasts as it has been shown that RANKL induces iNOS expression and NO generation as an autocrine negative feedback signal to limit osteoclastogenesis [469]. This inhibition of osteoclastogenesis induced by NO may be caused by NO-mediated inactivation of TGs via S-nitrosylation [206]. As expected, the induction of osteoclastogenesis with joint M-CSF and RANKL treatment significantly upregulated Nfatc1 (NFATc1), Acp5 (TRAP), Ctsk (CTSK), Dcstamp (DC- STAMP) and Atp6v0d2 (Atp6v0d2) expression (Figure 2) as well as resulted in a significant increase in TRAP activity in the culture supernatant on day 3 and onwards (Figure 3A). In our previous study, we have shown that TRAP+ mononuclear pre-osteoclasts are formed on day 2

96 and fusion occurs from day 3 [291]. Visualizing nuclei gave rise to the same results, i.e., cells were mononuclear on day 2 and became multinucleated on day 3. Visualizing F-actin showed that podosome clusters were visible on day 3, an expanding podosome ring on day 4, and a stable podosome belt on day 5 and 6 (Figure 3B). These changes represent the dynamic cytoskeletal organization that occur during the fusion stage [148].

Day-by-day evaluation of total TG activity during osteoclastogenesis and its regulation by M-CSF and RANKL In light of the above description of the differentiation process, we set to analyze TG activity during osteoclastogenesis using following approaches and activity assays. Total activity was assessed using 5-(biotinamido)pentylamine (bPA) which is a primary amine that incorporates to glutamine residues of TG substrates [470]. For analyzing individual activities of FXIII-A, TG2, and TG1, we used biotinylated “Hitomi peptides”, bF11, bT26, bK5, respectively. In their control peptides, Q-residues are replaced by N-residues which renders the peptide inactive toward TGs [211-213]. All reagents are detailed in Table I. All assays were done with endogenous substrates present in the cell extracts, i.e., no exogenous substrates, such as casein or fibronectin were added. All assays included Ca2+ and thiol-group reducing agent DTT which is required for activation of TGs [471]. Covalent incorporation of the bPA or biotinylated peptides to substrates was visualized by Western blotting and detection of biotin. Analysis of total TG activity with bPA during osteoclastogenesis showed high activity from day 0 to day 3 and absence of activity from day 4 to day 6 (Figure 4A). The examination of the effects of M-CSF and RANKL on the total TG activity showed that M-CSF at 200 ng/ml dramatically upregulated TG activity and the addition of RANKL downregulated it (Figure 4B). Consistently, the individual activities of FXIII-A, TG2, and TG1 also showed the same pattern. These are detailed below for each enzyme.

Day-by-day evaluation of FXIII-A activity, gene and protein expression levels during osteoclastogenesis We next analyzed the individual activities as well as protein and mRNA expression of TGs

97 during osteoclastogenesis. mRNA analyses of TGs treated with M-CSF alone were conducted as a reference and significance of expression levels between the days was only done for the joint treatment with M-CSF and RANKL. F13a1 mRNA expression was significantly and dramatically downregulated by the joint treatment with M-CSF and RANKL (Figure 5A). FXIII-A protein was also downregulated by the same treatment (Figure 5B). FXIII-A activity, assessed with bF11, had the same pattern, i.e., FXIII-A activity was highest from day 0 to day 2 and downregulated from day 3 to day 6 (Figure 5C). We also analyzed if the mRNA, protein expression and activities of TGs were regulated by M- CSF and RANKL. F13a1 mRNA expression was significantly upregulated by M-CSF at 100 ng/ml but downregulated by M-CSF at 200 ng/ml and dramatically and significantly downregulated by the addition of RANKL (Figure 6A). Consistently, increasing M-CSF concentration enhanced the level of FXIII-A protein (Figure 6B) and 200 ng/ml M-CSF induced a dramatic increase in FXIII-A activity (Figure 6C). RANKL resulted in a decrease in both protein and activity. Taken together, these data suggest that FXIII-A is involved in the M- CSF-mediated osteoclast precursor commitment phase.

Day-by-day evaluation of TG2 activity and gene and protein expression levels during osteoclastogenesis Analysis of Tgm2 mRNA expression showed its significant upregulation by the joint M-CSF and RANKL treatment which peaked on day 4 and then showed a significant downregulation from day 4 to day 6 (Figure 7A). This was also reflected in TG2 protein level which was upregulated from day 0 to day 3 and downregulated from day 4 to day 6 (Figure 7B). Interestingly, TG2 was detected mainly at molecular weights 75 kDa, 250 kDa, and above 250 kDa, plus two faint bands at 130 kDa and 180 kDa. While the identities of these bands remain unknown, it is possible that they represent TG2 complexes that arise from covalent complexation with itself [472] or other proteins [238]. These complexes increased with increasing M-CSF concentration (Figure 8B). Consistently, TG2 activity, assessed with bT26, was upregulated from day 0 to day 3 and downregulated from day 4 to day 6 (Figure 7C). Tgm2 mRNA expression was significantly and dose-dependently upregulated by the addition

98 of RANKL (Figure 8A). Protein analysis showed an upregulation of TG2 by M-CSF (Figure 8B) and a slight downregulation by the addition of RANKL (Figure 8B). M-CSF also increased TG2 activity, whereas addition of RANKL caused a decrease (Figure 8C).

Day-by-day evaluation of TG1 activity and gene and protein expression levels during osteoclastogenesis Tgm1 mRNA expression was upregulated by the joint treatment with M-CSF and RANKL which showed a significant increase on day 4 and significant decrease after that (Figure 9A). TG1 protein level was high from day 0 to day 3, followed by a decrease from day 4 to day 6 (Figure 9B). Consistently, TG1 activity, assessed with bK5, was high from day 0 to day 3 (Figure 9C), and decreased from day 4 to day 6, however, some residual activity was seen from day 4 to day 6 with substrates at very high molecular weight. Tgm1 mRNA expression was significantly upregulated by M-CSF and significantly downregulated by the addition of RANKL (Figure 10A). TG1 protein level did not change with increasing M-CSF concentration although activity did (Figure 10B, C). Addition of RANKL downregulated TG1 protein and activity (Figure 10B).

TG activity plays an important role in the differentiation and early fusion stages of osteoclastogenesis Based on above data, it is clear that total TG activity and FXIII-A, TG2 and TG1 activities are highest in the early phase of osteoclastogenesis, i.e., from day 0 to day 3 and then decreases dramatically from day 4 to 6 (Figure 4). In our previous study, we showed that TRAP+ mononuclear pre-osteoclasts are formed on day 2 and fusion occurs from day 3 onwards [291]. We also showed that the TG inhibitor NC9, when given to the cells during differentiation phase (day 1 and day 2) and fusion phase ( day 3 to day 5), suppress both processes [291]. In the light of the new findings, we aimed to redefine the role of TG activity in osteoclastogenesis, and now inhibited TG activity from day 1 to day 3 and compared that to TG inhibition from day 1 to day 6, or from day 4 to day 6 (Figure 11A). Osteoclastogenesis was assessed on day 6 with TRAP staining and TRAP activity in the culture supernatant. Figure 11B showed clearly that

99 inhibition of TG activity from day 1 to day 3 resulted in dramatic decrease in TRAP activity and that the effect was the same as inhibition from days 1 to day 6. Inhibition from days 4 to day 6 also showed an inhibitory effect on the TRAP activity, but to a lesser extent. Figure 11C showed a complete lack of large multinucleated osteoclasts with NC9 treatments from day 1 to day 6 and from day 1 to day 3, whereas abundant large osteoclasts were present with inhibition from day 4 to day 6. These data support our conclusion that TG activity plays a more important role in the differentiation and early fusion stages of osteoclastogenesis. In summary, the three TGs likely exert distinct functions at early stage of osteoclastogenesis, i.e., FXIII-A is likely involved in M-CSF-mediated precursor commitment phase, and both TG1 and TG2 likely play an important role in the differentiation and early fusion stages.

4.5 Discussion Osteoclastogenesis occurs in phases which involve precursor commitment, differentiation and fusion [16,465], with a dynamic cytoskeletal organization concomitant with the fusion phase [148]. Transglutaminases TG1, TG2, and FXIII-A have been linked to osteoclastogenesis in several recent studies, however, the current understanding is limited to the observations that FXIII-A null mice have defective osteoclastogenesis [289] and TG2 null mice have increased osteoclastogenesis [290]. In addition, TG1 is expressed in both wild type and TG2/FXIII-A double null mice, the latter one displaying an increased osteoclastogenesis [288]. It is also known that TG inhibitor NC9 blocks osteoclastogenesis in both wild type and TG2/FXIII-A double deficient cells [288]. To start understanding the separate roles of these TGs in osteoclastogenesis, our aim was to screen in detail their temporal expression and activities and to examine if they are regulated by the key osteoclastogenic factors, M-CSF and RANKL. In this study, we demonstrate that the three TGs have different expression and activity patterns, suggesting separate roles for the three enzymes in osteoclastogenesis. FXIII-A was first demonstrated present in monocytes and macrophages in 1985 [328,473] and studies have demonstrated that a transient intranuclear accumulation of FXIII-A occurs in the early phase of macrophage differentiation and that FXIII-A in the nuclei exhibited cross-linking activity [356]. In line with previous studies, our results showed M-CSF dramatically enhanced

100 the level of FXIII-A protein, strongly suggesting its involvement in the M-CSF-induced macrophage differentiation, i.e., the osteoclast precursor commitment. This may be part of the mechanism by which FXIII-A mediates regulation of osteoclastogenesis as demonstrated by the data from FXIII-A deficient mice [289]. It is however, important to note that the numbers and percentages of CD11b+ myeloid cells which gave rise to macrophages were similar in wild type and FXIII-A deficient mice [289]. In addition, when treated with M-CSF, both wild type and FXIII-A deficient cultures revealed similar numbers of macrophages [289]. Therefore, it seems that FXIII-A does not affect the macrophage differentiation but perhaps promotes osteoclastogenesis via regulating osteoclast specific molecules that are expressed during osteoclast precursor commitment stage, such as RANK [56]. The mRNA and protein of FXIII- A were dramatically downregulated by RANKL, suggesting that it is not involved in any RANKL-induced events. Based on the opposite mouse phenotypes of TG2- and FXIII-A-deficient mice, an opposite expression pattern for TG2 to FXIII-A would be expected [289,290], however, also TG2 was upregulated by M-CSF and downregulated by the addition of RANKL albeit not as dramatically as FXIII-A. Interestingly, TG2 was detected as monomer, but also as high molecular weight complexes throughout osteoclastogenesis. Since all the gels in this study have been run in denaturing conditions, these are covalently linked complexes. Studies have shown that in the presence of Ca2+, TG2 efficiently utilizes itself as a substrate, leading to the formation of covalently linked TG2 multimers [472]. It has been demonstrated that cytosolic Ca2+ oscillations occur during RANKL-mediated osteoclastogenesis [474]. Therefore, the distinct bands detected could be monomer, dimer, trimer and higher-order multimers of TG2. Interestingly, this self-multimerization is enhanced with M-CSF. In addition, several types of TG2-containing protein complexes have been identified on the surface of various cells [238]. For instance, TG2 can form a ternary complex with integrin β1 or β3 and fibronectin to promote cell adhesion and spreading of fibroblasts, and this function does not require its cross-linking activity [244]. In addition, TG2 is associated with integrin β1 and β3 in monocytes and promotes the adhesion and migration of monocytes on fibronectin [323]. Thus, it is also possible that TG2 is auto-crosslinked with integrin or fibronectin in osteoclasts. A recent study

101 also showed that TG2 inhibited excessive osteoclastogenesis, and TG2 knockdown activated NF-κB via enhancing the phosphorylation of IκB (inhibitor of κB) via the non-crosslinking function of TG2 [290]. However, previous studies showed that TG2 in microglia activated NF- κB via inducing the polymerization of IκBα via its cross-linking activity [398]. These functionalities may be part of the mechanisms by which TG2 modulates osteoclastogenesis. Our future studies will focus on identifying the actions of TG2 in osteoclastogenesis and the nature of the TG2 complexes may provide mechanistical insight into this. The involvement of TG1 in osteoclastogenesis has been a novel finding from our laboratory and our previous work suggests that it may have a promoting role for osteoclastogenesis [288]. Previous studies have demonstrated that the 106 kDa TG1 shows low activity, but two proteolytically processed forms, i.e., one 67 kDa fragment and a complex consisting of 67/33 kDa fragments, had higher activities than the 106 kDa form [269]. In addition, it has been demonstrated that cathepsin D (CTSD) plays an essential role in processing TG1 precursor to an enzymatic active 35 kDa form [271]. Interestingly, TG2 has been found to associate with CTSD in mouse embryonic fibroblasts, leading to depletion of CTSD [475]. Therefore, TG2 may also regulate the proteolytic processing of TG1 via CTSD. In addition to distinct expression and activity patterns, the enzymes also appear to have several substrates in the cells. “Hitomi peptides” crosslink to K-residues of substrates in the presence of specific TGs, whereas bPA crosslinks to Q-residues in the presence of any TG activity. bF11 had clear labeling of 75 kDa, 63 kDa, 45 kDa, 37 kDa, and 30 kDa as FXIII-A substrates. bT26 had labeled one major TG2 substrate at 30 kDa and some minor ones at 45 kDa, 50 kDa, and 85 kDa. bK5 in turn showed most specific and strong labeling at 30 kDa, 37 kDa, 45 kDa, 300 kDa and above 300 kDa. The smear seen in 200 ng/ml M-CSF treatment may be caused by cross-linking between the abundantly labeled substrates. The “Hitomi peptides” do not penetrate cell membrane and thus in situ assay to examine TG activity in cells, or identification of substrates in this manner, was not possible. Thus, cell-permeable, fluorescence-labeled “Hitomi peptides” could be extremely valuable tools for in situ substrate identification. bPA showed very similar temporal overall activity pattern during osteoclastogenesis as “Hitomi peptides”, but showed different substrate labeling pattern as expected. However, it is interesting

102 to observe that some molecular weights of the bPA labeled proteins are similar with those of the “Hitomi peptide”-labeled proteins, i.e., 63 kDa (TG2) and 45 kDa (TG1). Our future work aims at identifying all Q- and K-substrates via proteomics to be able to fully understand the functions of all TGs during osteoclastogenesis. It is also worthy of mention that activity assays with endogenous substrates versus exogenous substrates added to assay can result in very different outcomes. For example, in our previous study [291], we used the same culture system but measured the total TG activity with bPA and exogenous substrate fibronectin [291]. This resulted in a fairly high activity on day 6 in mature osteoclasts whereas with endogenous substrate there is essentially no activity. Thus, we think it is closer to actual conditions to use endogenous substrates as it reflects the whole TG realm of the cells. Furthermore, in situ analysis may expand the view of this realm as it may include many small molecule regulators such as Ca2+ oscillations and iNOS levels which showed dramatic rise during osteoclastogenesis and may cause NO-mediated inactivation of TGs via S-nitrosylation [206]. Assessing the activity of individual enzymes has been a great challenge in the TG field. The discovery of “Hitomi peptides” has provided the field a tremendous tools. Our work here further validates their usefulness and specificity in a system where three TGs are expressed and present in cells at the same time. The peptides all demonstrated distinct labeling from each other in a manner that followed the expression patterns of the enzymes. Moreover, we now are equipped with novel hypothesis for the functions of the three TGs in osteoclastogenesis.

4.6 Acknowledgments The authors thank Aida Kasaei-Roodsari for helpful assistance. This study was supported by grants to MTK from the Canadian Institutes of Health Research (CIHR) (MOP-119403, PJT- 162100). HS is supported by stipend from the Faculty of Dentistry of McGill University. MTK is members of the Fonds de Recherche – Santé (FRQ-S) Network for Oral and Bone Health Research.

103

FIGURE 1 4.7 Figures

****

D0 Bone marrow macrophage D6

M-CSF

**** *** ****

****

FIGURE 1 M-CSF maintains macrophage phenotype of BMMs. BMMs were treated with M-CSF for 6 days and the expression of macrophage markers was evaluated on each day. The pan-macrophage marker Adgre1 (F4/80) was upregulated by the treatment and peaked on day 6. M2 macrophage marker Arg1 (ARG1) was downregulated by the treatment and was the lowest at day 2, 4 and 6, whereas M1 macrophage marker Nos2 (iNOS) was upregulated by the treatment but the increase was only significant at day 6, suggesting that M-CSF treated BMMs likely adopt the M1 macrophage phenotype. Error bars represent SEM. n = 3. ***p < 0.001, ****p < 0.0001.

104 FIGURE 2

****

D0 Osteoclastogenesis D6

M-CSF + RANKL

*** ****

** ****

105 FIGURE 2 M-CSF and RANKL induce osteoclastogenesis from BMMs. BMMs were treated with M-CSF and RANKL for 6 days and the expression of osteoclast markers was evaluated on each day. M-CSF and RANKL treatment upregulated the expression of osteoclast differentiation markers Nfatc1 (NFATc1), Acp5 (TRAP) and Ctsk (CTSK), as well as osteoclast fusion markers Dcstamp (DC-STAMP) and Atp6v0d2 (Atp6v0d2), confirming the osteoclastogenesis induced by M-CSF and RANKL. Error bars represent SEM. n = 3. **p < 0.01, ***p < 0.001, ****p < 0.0001.

106 FIGURE 3

A B Day 1 Day 2 Day 3

**** Actin +Actin

Day 4 Day 5 Day 6 DAPI

FIGURE 3 M-CSF and RANKL enhance TRAP activity during osteoclastogenesis and induce the dynamic cytoskeletal organization during the fusion stage. BMMs were treated with M-CSF and RANKL for 6 days and the TRAP activities in the culture supernatant were measured and the nuclei as well as F-actin were stained on each day. (A) M-CSF plus RANKL treatment significantly increased the TRAP activity in the culture supernatant on day 3 and onwards. Error bars represent SEM. n = 3. ****p < 0.0001. (B) Visualizing nuclei with DAPI staining (blue) showed that cells were mononuclear on day 2 and became multinucleated on day 3, suggesting fusion starts on day 3. Visualizing actin with Alexa Fluor® 488-phalloidin (green) showed that podosome clusters were visible at day 3, an expanding podosome ring at day 4, and a stable podosome belt on day 5 and day 6, indicating the dynamic cytoskeleton organization during the fusion stage. Magnification bar represents 200 μm.

107 TABLE I

Reagent/ Sequence Assay peptide Conc. bPA 5-(biotinamido)pentylamine 500 μM Total TG activity bK5 Biotinyl-YEQHKLPSSWPF 150 μM TG1 activity bK5QN Biotinyl-YENHKLPSSWPF 150 μM TG1 control bT26 Biotinyl-HQSYVDPWMLDH 80 μM TG2 activity bT26QN Biotinyl-HNSYVDPWMLDH 80 μM TG2 control bF11 Biotinyl-DQMMLPWPAVAL 40 μM FXIII-A activity bF11QN Biotinyl-DNMMLPWPAVAL 40 μM FXIII-A control

TABLE I Illustration of all the reagents used in TG activity assay in this study.

108 activity upregulated with maintained BMMs Total incorporation days regulation FIGURE bPA TG and were assays /Western activity 4 by the at by Day treated M of the were M total - CSF - bPA - by highest was CSF blot - independently B A with day kDa kDa 100 150 250 150 250 (biotin) and 100 TG 45 63 63 45 approach high at evaluation increasing 200 concentration) RANKL activity Actin Biotin Actin Biotin from - M bPA to ng/ml 50 0 - CSF (ng/ml) . substrate Results day 100 0 . repeated (A) was M and of 1 0 - 200 Days (M CSF FIGURE 4 FIGURE total BMMs to downregulated for showed evaluated 2 proteins day twice + M 50 6 or TG - RANKL (ng/ml) 3 - CSF + CSF CSF (200 ng/ml) days with were 3 100 . that 4 activity but RANKL) in and increasing with treated 5 200 the the was the by 6 total cell during bPA absent

addition

total

with activity assay activity - TG / bPA

activity assay activity - TG / extracts bPA TG RANKL followed TG M osteoclastogenesis from activity - of CSF activity via RANKL day (while and Western was by 4 was RANKL to detection dramatically . M All day - evaluated CSF blotting the and 6 . for was TG (B) 109 its of 6 . FIGURE 5

****

B Days (M-CSF + RANKL) 0 1 2 3 4 5 6 kDa

80 Actin

Days (M-CSF + RANKL) bF11QN

0 0 1 2 3 4 5 6

bF11

/ / FXIII

Biotin

- A A

kDa - activityassay 75 63 45 37

30 Actin

110 FIGURE 5 Day-by-day evaluation of mRNA, protein expression and activity of FXIII-A during osteoclastogenesis. BMMs were treated with M-CSF and RANKL for 6 days and the mRNA and protein expressions and activity of FXIII-A were evaluated on each day. (A) F13a1 mRNA expression was significantly and dramatically downregulated by the joint treatment with M-CSF and RANKL. Error bars represent SEM. n = 3. ****p < 0.0001. (B) FXIII-A protein was also downregulated by the joint treatment with M-CSF and RANKL. (C) FXIII-A activity was evaluated with bF11 and bF11QN using a Western blot detection of biotinylated- peptide incorporation to substrates in cell extracts. Results showed that FXIII-A activity was highest from day 0 to day 2 and downregulated from day 3 to day 6. All the protein levels and TG activity assays were independently repeated twice.

111 FIGURE 6

A **** ****

****

50 100 200 50 100 200 M-CSF (ng/ml) RANKL (ng/ml) +M-CSF (200 ng/ml)

B M-CSF (200 ng/ml) M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200 kDa

80 Actin

C M-CSF (200 ng/ml) M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200

bF11

Biotin

/ / FXIII

- Aactivity assay kDa 50 45

30 Actin

112 FIGURE 6 Regulation of mRNA, protein expression and activity of FXIII-A by M-CSF and RANKL. BMMs were treated with increasing M-CSF or with increasing RANKL (while M-CSF was maintained at the highest concentration) for 6 days and the mRNA and protein expressions and activity of FXIII-A were evaluated on day 6. (A) F13a1 mRNA expression was significantly upregulated by M-CSF at 100 ng/ml but downregulated by M-CSF at 200 ng/ml and dramatically and significantly downregulated by the addition of RANKL. Error bars represent SEM. n = 3. *p < 0.05, ****p < 0.0001. (B) Increasing M- CSF concentration enhanced the level of FXIII-A protein which was dramatically downregulated by the addition of RANKL. (C) Activity assay with bF11, that involved Western blot detection of biotinylated-peptide incorporation to substrates in cell extracts, showed dramatic upregulation by 200 ng/ml M-CSF. Addition of RANKL has a dramatic downregulation effect on FXIII-A activity. All the protein levels and TG activity assays were independently repeated twice.

113 FIGURE 7

***

****

B Days (M-CSF + RANKL) kDa 0 1 2 3 4 5 6 250 180 130

75 Actin

Days (M-CSF + RANKL) bT26QN

0 0 1 2 3 4 5 6

bT26

Biotin

/ TG2 / -

kDa activity assay 85 50 45

30 Actin

114 FIGURE 7 Day-by-day evaluation of mRNA, protein expression and activity of TG2 during osteoclastogenesis. BMMs were treated with M-CSF and RANKL for 6 days and the mRNA and protein expressions and activity of TG2 were evaluated on each day. (A) Tgm2 mRNA expression was significantly upregulated by the joint M-CSF and RANKL treatment which peaked on day 4 and then showed a significant downregulation from day 4 to day 6. Error bars represent SEM. n = 3. ***p < 0.001, ****p < 0.0001. (B) TG2 protein was upregulated from day 0 to day 3 and downregulated from day 4 to day 6. TG2 was detected mainly at molecular weights 75 kDa, 250 kDa, and above 250 kDa, plus two faint bands at 130 kDa and 180 kDa, suggesting covalent complexation with itself or other proteins. (C)

TG2 activity assessed with bT26 and bT26QN in assay where the biotinylated-peptide incorporation to substrates in cell extracts was detected Western blot. TG2 activity was upregulated from day 0 to day 3 and downregulated from day 4 to day 6. All the protein levels and TG activity assays were independently repeated twice.

115 FIGURE 8

A ** ** ***

50 100 200 50 100 200 M-CSF (ng/ml) RANKL (ng/ml) +M-CSF (200 ng/ml)

M-CSF (200 ng/ml) B M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200 kDa 250 180 130

75 Actin

C M-CSF (200 ng/ml) M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200

bT26

Biotin / TG2 /

kDa

- activity assay 85 63 50 45

30 Actin

116 FIGURE 8 Regulation of mRNA, protein expression and activity of TG2 by M-CSF and RANKL. BMMs were treated with increasing M-CSF or with increasing RANKL (while M-CSF was maintained at the highest concentration) for 6 days and the mRNA and protein expressions and activity of TG2 were evaluated on day 6. (A) Tgm2 mRNA expression was significantly and dose-dependently upregulated by the addition of RANKL. Error bars represent SEM. n = 3. **p < 0.01, ***p < 0.001. (B) TG2 protein level was upregulated by M-CSF and slightly downregulated by addition of RANKL. (C) TG2 activity assessed via Western blot detection of biotin of bT26 that was covalently incorporated to substrate peptides in cell extracts during the assay. TG2 activity was upregulated by M-CSF but dramatically decreased by addition of RANKL. All the protein levels and TG activity assays were independently repeated twice.

117 FIGURE 9

**** ***

****

B Days (M-CSF + RANKL)

kDa 0 1 2 3 4 5 6

100 Actin

Days (M-CSF + RANKL) bK5QN

0 0 1 2 3 4 5 6 bK5 kDa

300 TG1 /

activity assay Biotin

45 37

30 Actin

118 FIGURE 9 Day-by-day evaluation of mRNA, protein expression and activity of TG1 during osteoclastogenesis. BMMs were treated with M-CSF and RANKL for 6 days and the mRNA and protein expressions and activity of TG1 were evaluated on each day. (A) Tgm1 mRNA expression was upregulated by the joint treatment with M-CSF and RANKL which showed a significant increase on day 4 and significant decrease after that. Error bars represent SEM. n = 3. ***p < 0.001, ****p < 0.0001. (B) TG1 protein level was high from day 0 to day 3, followed by a decrease from day 4 to day 6. (C) TG1 activity assessed with bK5 and bK5QN via detection of biotin of the peptides after its covalent incorporation to substrate protein in cell extracts. TG1 activity was high from day 0 to day 3, and then decreased from day 4 to day 6, however, some residual activity was seen from day 4 to day 6 with substrates at very high molecular weight. All the protein levels and TG activity assays were independently repeated twice.

119 FIGURE 10 A ** **

50 100 200 50 100 200 M-CSF (ng/ml) RANKL (ng/ml) + M-CSF (200 ng/ml)

B M-CSF (200 ng/ml) M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200 kDa

100 Actin

C M-CSF (200 ng/ml) M-CSF (ng/ml) + RANKL (ng/ml)

50 100 200 50 100 200

bK5

/ TG1 /

Biotin

- activity assay

kDa 45

30 Actin

120 FIGURE 10 Regulation of mRNA, protein expression and activity of TG1 by M-CSF and RANKL. BMMs were treated with increasing M-CSF or with increasing RANKL (while M-CSF was maintained at the highest concentration) for 6 days and the mRNA and protein expressions and activity of TG1 were evaluated on day 6. (A) Tgm1 mRNA expression was significantly upregulated by M-CSF and significantly downregulated by addition of RANKL. Error bars represent SEM. n = 3. **p < 0.01. (B) TG1 protein was maintained at similar levels with increasing M-CSF, but was downregulated by addition of RANKL. (C) TG1 activity assessed with bK5 followed by Western blot detection of biotin of the peptides that were incorporated to substrate proteins in cell extracts. TG1 activity by was upregulated by M-CSF and downregulated by addition of RANKL. All the protein levels and TG activity assays were independently repeated twice.

121 FIGURE 11

NS A Osteoclastogenesis B **** M-CSF + RANKL **** **** **** Differentiation Fusion D0 D3 D6

NC9

NC9 NC9 DMSO NC9 DMSO NC9 DMSO NC9 1-6 1-6 1-3 1-3 4-6 4-6 C DMSO 1-6 DMSO 1-3 DMSO 4-6

NC9 1-6 NC9 1-3 NC9 4-6

FIGURE 11 TG activity plays an important role in the differentiation and early fusion stages of osteoclastogenesis. (A) The experimental design for the TG activity inhibitor, NC9 treatment. Inhibitor was given to the cells either from day 1 to day 6, from day 1 to day 3, or from day 4 to day 6. DMSO is used as a vehicle. (B) TRAP activities in the culture supernatant on day 6 showed that TG activity played a more critical role from day 1 to day 3, which represents the differentiation and early fusion stages. Error bars represent SEM. n = 3. ****p < 0.0001, NS: not significant. (C) TRAP staining of osteoclasts at day 6 formed with different NC9 treatments showed complete lack of large multinucleated osteoclasts when NC9 was given from day 1 to day 6, or from day 1 to day 3, which represents the differentiation and early fusion stages. Magnification bar represents 500 μm.

122 SUPPLEMENTAL FIGURE S1

**** NS

D0 Osteoclastogenesis D6

M-CSF + RANKL

NS ****

FIGURE S1 M-CSF and RANKL induce the similar expression pattern of macrophage markers in BMMs to M-CSF. BMMs were treated with M-CSF and RANKL for 6 days and the expression of macrophage markers was evaluated on each day. The pan-macrophage marker Adgre1 (F4/80) was upregulated by the treatment and peaked on day 4. M2 macrophage marker Arg1 (ARG1) was downregulated by the treatment with the lowest expression on day 2 and 4, whereas M1 macrophage marker Nos2 (iNOS) was upregulated dramatically by the treatment and peaked on day 6. Error bars represent SEM. n = 3. *p < 0.05, ****p < 0.0001, NS: not significant.

123 Chapter 5 – Conclusions, discussion and future directions

5.1 Original contributions  Transglutaminase activity regulates osteoclastogenesis.  Three transglutaminases are found in osteoclasts, TG2, FXIII-A and TG1.  Among the four different TG activity inhibitors tested, NC9 is the most potent one to inhibit osteoclastogenesis.  NC9 inhibits differentiation, migration, and fusion of pre-osteoclasts as well as the podosome belt formation and resorption of osteoclasts.  NC9 increases expression level of RhoA and the inhibitory effect of NC9 on osteoclastogenesis and podosome belt formation can be completely reversed with a Rho-family inhibitor, suggesting that TG activity regulates actin dynamics in osteoclasts.  General TG activity, and the activities of individual TGs, are highest in the differentiation and early fusion phases and then decrease dramatically.  Activities of all TGs are upregulated by M-CSF and downregulated by addition of RANKL.  FXIII-A mRNA and protein expression levels are dramatically downregulated by RANKL, suggesting its involvement in M-CSF-mediated precursor commitment phase.  TG1 and TG2 proteins are present throughout the osteoclastogenesis, suggesting their functions in both differentiation and fusion.  Hitomi peptides are highly specific tools for detection of distinct TGs in a system where multiple TGs are present.

5.2 Conclusions and discussion Total TG activity: Treatment with TG activity inhibitor NC9 shows that TG activity is required for osteoclastogenesis and affects both differentiation, migration, and fusion of pre- osteoclasts as well as the podosome belt formation and resorption of osteoclasts. Chemical inhibition of TG activity increases expression level of RhoA in osteoclasts. Rho-family inhibitor can reverse the inhibitory effect of TG inhibitor on osteoclastogenesis and podosome belt formation, suggesting that TG activity regulates actin dynamics in osteoclasts. Total TG activity is high from day 0 to day 3 (differentiation and early fusion phases) and then decreases dramatically from day 4 to day 6. In addition, NC9 treatments from day 1 to day 6 and from

124 day 1 to day 3 result in a complete lack of large multinucleated osteoclasts, whereas abundant large osteoclasts are present with inhibition from day 4 to day 6. Therefore, TG activity plays a more important role in the differentiation and early fusion stages of osteoclastogenesis.

FXIII-A: FXIII-A is upregulated by M-CSF and downregulated immediately by RANKL treatment, suggesting its involvement in osteoclast precursor commitment stage without participating in the RANKL-induced osteoclast differentiation. FXIII-A deficiency showed decreased osteoclastogenesis in vivo and in vitro [289]. However, the numbers and percentages of CD11b+ myeloid cells which gave rise to macrophages were similar in wild type and FXIII- A deficient mice [289]. In addition, when treated with M-CSF, both wild type and FXIII-A deficient cultures revealed similar numbers of macrophages [289]. Therefore, FXIII-A does not affect the macrophage differentiation but promotes osteoclastogenesis via regulating osteoclast specific molecules that are expressed during osteoclast precursor commitment stage, such as M-CSF induced RANK expression [56].

TG2: TG2 is upregulated by M-CSF and downregulated by the addition of RANKL albeit not as dramatically as FXIII-A. Interestingly, TG2 is detected as monomer, but also as high molecular weight complexes throughout osteoclastogenesis. Since all the gels in this study have been run in denaturing conditions, these are covalently linked complexes. Studies have shown that in the presence of Ca2+, TG2 efficiently utilizes itself as a substrate, leading to the formation of covalently linked TG2 multimers [472]. It has been demonstrated that cytosolic Ca2+ oscillations occur during RANKL-mediated osteoclastogenesis [474]. Therefore, the distinct bands detected could be monomer, dimer, trimer and higher-order multimers of TG2. Interestingly, this self-multimerization is enhanced with M-CSF. In addition, several types of TG2-containing protein complexes have been identified on the surface of various cells [238]. For instance, TG2 can form a ternary complex with integrin β1 or β3 and fibronectin to promote cell adhesion and spreading of fibroblasts, and this function does not require its cross-linking activity [244]. In addition, TG2 is associated with integrin β1 and β3 in monocytes and promotes the adhesion and migration of monocytes on fibronectin [323]. Thus, it is also possible that TG2 is auto-crosslinked with integrin or fibronectin in osteoclasts. A recent study also showed that TG2 inhibited excessive osteoclastogenesis, and TG2 knockdown activated NF-κB via enhancing the phosphorylation of IκB via the non-crosslinking function of TG2 [290]. However, previous studies showed that TG2 in microglia activated NF-κB via inducing

125 the polymerization of IκBα via its cross-linking activity [398]. These functionalities may be part of the mechanisms by which TG2 modulates osteoclastogenesis.

TG1: Our Tgm2-/-;F13a1-/- work suggests that it may have a promoting role for osteoclastogenesis [288]. Previous studies have demonstrated that the 106 kDa TG1 shows low activity, but two proteolytically processed forms, i.e., one 67 kDa fragment and a complex consisting of 67/33 kDa fragments, have higher activities than the 106 kDa form [269]. In addition, it has been demonstrated that CTSD plays an essential role in processing TG1 precursor to an enzymatic active 35 kDa form [271]. Interestingly, TG2 has been found to associate with CTSD in mouse embryonic fibroblasts, leading to depletion of CTSD [475]. Therefore, TG2 may also regulate the proteolytic processing of TG1 via CTSD.

5.3 Future work • Generate osteoclast-specific Tgm1 knockout mice and examine the roles of TG1 during osteoclastogenesis in vivo and in vitro. • Identify all the TG substrates via using biotin-labelled substrates with streptavidin- coated magnetic beads and proteomics. • Examine the effect of absence of one TG on the expression of the other two TGs in osteoclasts. • Develop the antibody against NC9 (against its dansyl group) and identify the TGs inhibited by NC9 in osteoclasts and the cellular location where inhibition occurs. • Explore the role of FXIII-A in the M-CSF-mediated precursor commitment phase. • Identify the components of TG2-containing complexes in osteoclasts. • Detect the level of CTSD and measure the TG1 activity in the Tgm2-/- or Tgm2- /-;F13a1-/- osteoclasts.

In summary, this thesis shows that TG activity is required for osteoclastogenesis and regulates this process via affecting actin dynamics. Expression and activities of TGs are regulated by M- CSF and RANKL and the three TGs likely exert distinct functions at different stages of osteoclastogenesis (Figure XI).

126

FIGURE XI

Bone marrow cells (Day -3) Commitment: FXIII-A

Bone marrow-derived macrophages (Day 0) Differentiation: TG1, TG2

TRAP+ pre-osteoclasts (Day 2)

Multinucleated osteoclasts (Day 3, podosome cluster)

Differentiation, migration, fusion, and podosome belt formation: TG activity Multinucleated osteoclasts (Day 4, podosome ring) Fusion: TG1, TG2

Multinucleated osteoclasts (Day 5, podosome belt)

Multinucleated osteoclasts (Day 6, podosome belt)

FIGURE XI The role of transglutaminases in osteoclastogenesis at different stages. TG activity regulates differentiation, migration, and fusion of pre-osteoclasts as well as the podosome belt formation and resorption of osteoclasts. TG activity plays a more important role in the differentiation and early fusion stages of osteoclastogenesis. FXIII-A mRNA and protein expression levels are dramatically downregulated by RANKL, suggesting its involvement in M- CSF-mediated precursor commitment phase. TG1 and TG2 proteins are present throughout the osteoclastogenesis, suggesting their functions in both differentiation and fusion. Therefore, the three TGs likely exert distinct functions at different stages of osteoclastogenesis.

127 REFERENCES

1. Florencio-Silva, R.; Sasso, G.R.; Sasso-Cerri, E.; Simoes, M.J.; Cerri, P.S. Biology of bone tissue: Structure, function, and factors that influence bone cells. BioMed research international 2015, 2015, 421746. 2. Feng, X.; McDonald, J.M. Disorders of bone remodeling. Annual review of pathology 2011, 6, 121-145. 3. Raggatt, L.J.; Partridge, N.C. Cellular and molecular mechanisms of bone remodeling. The Journal of biological chemistry 2010, 285, 25103-25108. 4. Rutkovskiy, A.; Stenslokken, K.O.; Vaage, I.J. Osteoblast differentiation at a glance. Medical science monitor basic research 2016, 22, 95-106. 5. Long, F. Building strong bones: Molecular regulation of the osteoblast lineage. Nature reviews. Molecular cell biology 2011, 13, 27-38. 6. Mackie, E.J. Osteoblasts: Novel roles in orchestration of skeletal architecture. The international journal of biochemistry & cell biology 2003, 35, 1301-1305. 7. Orimo, H. The mechanism of mineralization and the role of alkaline phosphatase in health and disease. Journal of Nippon Medical School = Nippon Ika Daigaku zasshi 2010, 77, 4-12. 8. Franz-Odendaal, T.A.; Hall, B.K.; Witten, P.E. Buried alive: How osteoblasts become osteocytes. Developmental dynamics : an official publication of the American Association of Anatomists 2006, 235, 176-190. 9. Heino, T.J.; Hentunen, T.A. Differentiation of osteoblasts and osteocytes from mesenchymal stem cells. Current stem cell research & therapy 2008, 3, 131-145. 10. Bellido, T. Osteocyte-driven bone remodeling. Calcified tissue international 2014, 94, 25-34. 11. Bonewald, L.F. The amazing osteocyte. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2011, 26, 229-238. 12. Delaisse, J.M. The reversal phase of the bone-remodeling cycle: Cellular prerequisites for coupling resorption and formation. BoneKEy reports 2014, 3, 561. 13. Everts, V.; Delaisse, J.M.; Korper, W.; Jansen, D.C.; Tigchelaar-Gutter, W.; Saftig, P.; Beertsen, W. The bone lining cell: Its role in cleaning howship's lacunae and initiating bone formation. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2002, 17, 77-90. 14. Streicher, C.; Heyny, A.; Andrukhova, O.; Haigl, B.; Slavic, S.; Schuler, C.; Kollmann, K.; Kantner, I.; Sexl, V.; Kleiter, M., et al. Estrogen regulates bone turnover by targeting rankl expression in bone lining cells. Scientific reports 2017, 7, 6460. 15. Ikeda, K.; Takeshita, S. The role of osteoclast differentiation and function in skeletal homeostasis. Journal of biochemistry 2016, 159, 1-8. 16. Yavropoulou, M.P.; Yovos, J.G. Osteoclastogenesis--current knowledge and future perspectives. Journal of musculoskeletal & neuronal interactions 2008, 8, 204-216. 17. Seita, J.; Weissman, I.L. Hematopoietic stem cell: Self-renewal versus differentiation. Wiley interdisciplinary reviews. Systems biology and medicine 2010, 2, 640-653. 18. Jacobsen, S.E.W.; Nerlov, C. Haematopoiesis in the era of advanced single-cell technologies. Nature cell biology 2019, 21, 2-8. 19. Auffray, C.; Sieweke, M.H.; Geissmann, F. Blood monocytes: Development, heterogeneity, and relationship with dendritic cells. Annual review of immunology 2009, 27, 669-692.

128

20. Ginhoux, F.; Jung, S. Monocytes and macrophages: Developmental pathways and tissue homeostasis. Nature reviews. Immunology 2014, 14, 392-404. 21. Sprangers, S.; de Vries, T.J.; Everts, V. Monocyte heterogeneity: Consequences for monocyte-derived immune cells. Journal of immunology research 2016, 2016, 1475435. 22. Mildner, A.; Marinkovic, G.; Jung, S. Murine monocytes: Origins, subsets, fates, and functions. Microbiology spectrum 2016, 4. 23. Dai, X.M.; Ryan, G.R.; Hapel, A.J.; Dominguez, M.G.; Russell, R.G.; Kapp, S.; Sylvestre, V.; Stanley, E.R. Targeted disruption of the mouse colony-stimulating factor 1 receptor gene results in osteopetrosis, mononuclear phagocyte deficiency, increased primitive progenitor cell frequencies, and reproductive defects. Blood 2002, 99, 111-120. 24. Cecchini, M.G.; Dominguez, M.G.; Mocci, S.; Wetterwald, A.; Felix, R.; Fleisch, H.; Chisholm, O.; Hofstetter, W.; Pollard, J.W.; Stanley, E.R. Role of colony stimulating factor-1 in the establishment and regulation of tissue macrophages during postnatal development of the mouse. Development (Cambridge, England) 1994, 120, 1357-1372. 25. Wiktor-Jedrzejczak, W.; Gordon, S. Cytokine regulation of the macrophage (m phi) system studied using the colony stimulating factor-1-deficient op/op mouse. Physiological reviews 1996, 76, 927-947. 26. Endele, M.; Loeffler, D.; Kokkaliaris, K.D.; Hilsenbeck, O.; Skylaki, S.; Hoppe, P.S.; Schambach, A.; Stanley, E.R.; Schroeder, T. Csf-1-induced src signaling can instruct monocytic lineage choice. Blood 2017, 129, 1691-1701. 27. Varol, C.; Mildner, A.; Jung, S. Macrophages: Development and tissue specialization. Annual review of immunology 2015, 33, 643-675. 28. Wynn, T.A.; Chawla, A.; Pollard, J.W. Macrophage biology in development, homeostasis and disease. Nature 2013, 496, 445-455. 29. Italiani, P.; Boraschi, D. From monocytes to m1/m2 macrophages: Phenotypical vs. Functional differentiation. Frontiers in immunology 2014, 5, 514. 30. Martinez, F.O.; Sica, A.; Mantovani, A.; Locati, M. Macrophage activation and polarization. Frontiers in bioscience : a journal and virtual library 2008, 13, 453-461. 31. Zhang, Y.H.; He, M.; Wang, Y.; Liao, A.H. Modulators of the balance between m1 and m2 macrophages during pregnancy. Frontiers in immunology 2017, 8, 120. 32. Schrader, J.W.; Moyer, C.; Ziltener, H.J.; Reinisch, C.L. Release of the cytokines colony-stimulating factor-1, granulocyte-macrophage colony-stimulating factor, and il-6 by cloned murine vascular smooth muscle cells. Journal of immunology (Baltimore, Md. : 1950) 1991, 146, 3799-3808. 33. Clinton, S.K.; Underwood, R.; Hayes, L.; Sherman, M.L.; Kufe, D.W.; Libby, P. Macrophage colony- stimulating factor gene expression in vascular cells and in experimental and human atherosclerosis. The American journal of pathology 1992, 140, 301-316. 34. Tsukui, T.; Kikuchi, K.; Mabuchi, A.; Sudo, T.; Sakamoto, T.; Sato, N.; Tsuneoka, K.; Shikita, M.; Aida, T.; Asano, G., et al. Production of macrophage colony-stimulating factor by adult murine parenchymal liver cells (hepatocytes). Journal of leukocyte biology 1992, 52, 383-389. 35. Fibbe, W.E.; Van Damme, J.; Billiau, A.; Duinkerken, N.; Lurvink, E.; Ralph, P.; Altrock, B.W.; Kaushansky, K.; Willemze, R.; Falkenburg, J.H. Human fibroblasts produce granulocyte-csf, macrophage-csf, and granulocyte-macrophage-csf following stimulation by interleukin-1 and poly(ri).Poly(rc). Blood 1988, 72, 860-866. 36. Takahashi, M.; Hong, Y.M.; Yasuda, S.; Takano, M.; Kawai, K.; Nakai, S.; Hirai, Y. Macrophage colony- stimulating factor is produced by human t lymphoblastoid cell line, cem-on: Identification by amino-

129

terminal amino acid sequence analysis. Biochemical and biophysical research communications 1988, 152, 1401-1409. 37. Naparstek, E.; Donnelly, T.; Shadduck, R.K.; Waheed, A.; Wagner, K.; Kase, K.; Greenberger, J.S. Persistent production of colony-stimulating factor (csf-1) by cloned bone marrow stromal cell line d2xrii after x-irradiation. Journal of cellular physiology 1986, 126, 407-413. 38. Elford, P.R.; Felix, R.; Cecchini, M.; Trechsel, U.; Fleisch, H. Murine osteoblastlike cells and the osteogenic cell mc3t3-e1 release a macrophage colony-stimulating activity in culture. Calcified tissue international 1987, 41, 151-156. 39. Zhao, S.; Zhang, Y.K.; Harris, S.; Ahuja, S.S.; Bonewald, L.F. Mlo-y4 osteocyte-like cells support osteoclast formation and activation. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2002, 17, 2068-2079. 40. Yu, W.; Chen, J.; Xiong, Y.; Pixley, F.J.; Dai, X.M.; Yeung, Y.G.; Stanley, E.R. Csf-1 receptor structure/function in maccsf1r-/- macrophages: Regulation of proliferation, differentiation, and morphology. Journal of leukocyte biology 2008, 84, 852-863. 41. Hamilton, J.A. Csf-1 signal transduction. Journal of leukocyte biology 1997, 62, 145-155. 42. Joos, H.; Trouliaris, S.; Helftenbein, G.; Niemann, H.; Tamura, T. Tyrosine phosphorylation of the juxtamembrane domain of the v-fms oncogene product is required for its association with a 55-kda protein. The Journal of biological chemistry 1996, 271, 24476-24481. 43. Bourette, R.P.; Rohrschneider, L.R. Early events in m-csf receptor signaling. Growth factors (Chur, Switzerland) 2000, 17, 155-166. 44. Mancini, A.; Niedenthal, R.; Joos, H.; Koch, A.; Trouliaris, S.; Niemann, H.; Tamura, T. Identification of a second grb2 binding site in the v-fms tyrosine kinase. Oncogene 1997, 15, 1565-1572. 45. Wilhelmsen, K.; Burkhalter, S.; van der Geer, P. C-cbl binds the csf-1 receptor at tyrosine 973, a novel phosphorylation site in the receptor's carboxy-terminus. Oncogene 2002, 21, 1079-1089. 46. Feng, X.; Teitelbaum, S.L. Osteoclasts: New insights. Bone research 2013, 1, 11-26. 47. Yu, W.; Chen, J.; Xiong, Y.; Pixley, F.J.; Yeung, Y.G.; Stanley, E.R. Macrophage proliferation is regulated through csf-1 receptor tyrosines 544, 559, and 807. The Journal of biological chemistry 2012, 287, 13694-13704. 48. Alonso, G.; Koegl, M.; Mazurenko, N.; Courtneidge, S.A. Sequence requirements for binding of src family tyrosine kinases to activated growth factor receptors. The Journal of biological chemistry 1995, 270, 9840-9848. 49. Lee, A.W.; States, D.J. Both src-dependent and -independent mechanisms mediate phosphatidylinositol 3-kinase regulation of colony-stimulating factor 1-activated mitogen-activated protein kinases in myeloid progenitors. Molecular and cellular biology 2000, 20, 6779-6798. 50. Xiong, Y.; Song, D.; Cai, Y.; Yu, W.; Yeung, Y.G.; Stanley, E.R. A csf-1 receptor phosphotyrosine 559 signaling pathway regulates receptor ubiquitination and tyrosine phosphorylation. The Journal of biological chemistry 2011, 286, 952-960. 51. Bourette, R.P.; Myles, G.M.; Choi, J.L.; Rohrschneider, L.R. Sequential activation of phoshatidylinositol 3-kinase and phospholipase c-gamma2 by the m-csf receptor is necessary for differentiation signaling. The EMBO journal 1997, 16, 5880-5893. 52. Xaus, J.; Comalada, M.; Valledor, A.F.; Cardo, M.; Herrero, C.; Soler, C.; Lloberas, J.; Celada, A. Molecular mechanisms involved in macrophage survival, proliferation, activation or apoptosis. Immunobiology 2001, 204, 543-550. 53. Monick, M.M.; Mallampalli, R.K.; Bradford, M.; McCoy, D.; Gross, T.J.; Flaherty, D.M.; Powers, L.S.;

130

Cameron, K.; Kelly, S.; Merrill, A.H., Jr., et al. Cooperative prosurvival activity by erk and akt in human alveolar macrophages is dependent on high levels of acid ceramidase activity. Journal of immunology (Baltimore, Md. : 1950) 2004, 173, 123-135. 54. Busca, A.; Saxena, M.; Iqbal, S.; Angel, J.; Kumar, A. Pi3k/akt regulates survival during differentiation of human macrophages by maintaining nf-kappab-dependent expression of antiapoptotic bcl-xl. Journal of leukocyte biology 2014, 96, 1011-1022. 55. Richardson, E.T.; Shukla, S.; Nagy, N.; Boom, W.H.; Beck, R.C.; Zhou, L.; Landreth, G.E.; Harding, C.V. Erk signaling is essential for macrophage development. PloS one 2015, 10, e0140064. 56. Arai, A.; Mizoguchi, T.; Harada, S.; Kobayashi, Y.; Nakamichi, Y.; Yasuda, H.; Penninger, J.M.; Yamada, K.; Udagawa, N.; Takahashi, N. Fos plays an essential role in the upregulation of rank expression in osteoclast precursors within the bone microenvironment. Journal of cell science 2012, 125, 2910-2917. 57. Felix, R.; Cecchini, M.G.; Hofstetter, W.; Elford, P.R.; Stutzer, A.; Fleisch, H. Impairment of macrophage colony-stimulating factor production and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 1990, 5, 781-789. 58. Yoshida, H.; Hayashi, S.; Kunisada, T.; Ogawa, M.; Nishikawa, S.; Okamura, H.; Sudo, T.; Shultz, L.D.; Nishikawa, S. The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature 1990, 345, 442-444. 59. Wojtowicz, A.; Dziedzic-Goclawska, A.; Kaminski, A.; Stachowicz, W.; Wojtowicz, K.; Marks, S.C., Jr.; Yamauchi, M. Alteration of mineral crystallinity and collagen cross-linking of bones in osteopetrotic toothless (tl/tl) rats and their improvement after treatment with colony stimulating factor-1. Bone 1997, 20, 127-132. 60. Lin, H.; Lee, E.; Hestir, K.; Leo, C.; Huang, M.; Bosch, E.; Halenbeck, R.; Wu, G.; Zhou, A.; Behrens, D., et al. Discovery of a cytokine and its receptor by functional screening of the extracellular proteome. Science (New York, N.Y.) 2008, 320, 807-811. 61. Wei, S.; Nandi, S.; Chitu, V.; Yeung, Y.G.; Yu, W.; Huang, M.; Williams, L.T.; Lin, H.; Stanley, E.R. Functional overlap but differential expression of csf-1 and il-34 in their csf-1 receptor-mediated regulation of myeloid cells. Journal of leukocyte biology 2010, 88, 495-505. 62. Hwang, S.J.; Choi, B.; Kang, S.S.; Chang, J.H.; Kim, Y.G.; Chung, Y.H.; Sohn, D.H.; So, M.W.; Lee, C.K.; Robinson, W.H., et al. Interleukin-34 produced by human fibroblast-like synovial cells in rheumatoid arthritis supports osteoclastogenesis. Arthritis research & therapy 2012, 14, R14. 63. Nelson, C.A.; Warren, J.T.; Wang, M.W.; Teitelbaum, S.L.; Fremont, D.H. Rankl employs distinct binding modes to engage rank and the osteoprotegerin decoy receptor. Structure (London, England : 1993) 2012, 20, 1971-1982. 64. Hikita, A.; Yana, I.; Wakeyama, H.; Nakamura, M.; Kadono, Y.; Oshima, Y.; Nakamura, K.; Seiki, M.; Tanaka, S. Negative regulation of osteoclastogenesis by ectodomain shedding of receptor activator of nf- kappab ligand. The Journal of biological chemistry 2006, 281, 36846-36855. 65. Ikeda, T.; Kasai, M.; Utsuyama, M.; Hirokawa, K. Determination of three isoforms of the receptor activator of nuclear factor-kappab ligand and their differential expression in bone and thymus. Endocrinology 2001, 142, 1419-1426. 66. Douni, E.; Rinotas, V.; Makrinou, E.; Zwerina, J.; Penninger, J.M.; Eliopoulos, E.; Schett, G.; Kollias, G. A rankl g278r mutation causing osteopetrosis identifies a functional amino acid essential for trimer assembly in rankl and tnf. Human molecular genetics 2012, 21, 784-798. 67. Yasuda, H.; Shima, N.; Nakagawa, N.; Yamaguchi, K.; Kinosaki, M.; Mochizuki, S.; Tomoyasu, A.;

131

Yano, K.; Goto, M.; Murakami, A., et al. Osteoclast differentiation factor is a ligand for osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical to trance/rankl. Proceedings of the National Academy of Sciences of the United States of America 1998, 95, 3597-3602. 68. Lacey, D.L.; Timms, E.; Tan, H.L.; Kelley, M.J.; Dunstan, C.R.; Burgess, T.; Elliott, R.; Colombero, A.; Elliott, G.; Scully, S., et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 1998, 93, 165-176. 69. Boyle, W.J.; Simonet, W.S.; Lacey, D.L. Osteoclast differentiation and activation. Nature 2003, 423, 337- 342. 70. Xiong, J.; Piemontese, M.; Onal, M.; Campbell, J.; Goellner, J.J.; Dusevich, V.; Bonewald, L.; Manolagas, S.C.; O'Brien, C.A. Osteocytes, not osteoblasts or lining cells, are the main source of the rankl required for osteoclast formation in remodeling bone. PloS one 2015, 10, e0138189. 71. Anderson, D.M.; Maraskovsky, E.; Billingsley, W.L.; Dougall, W.C.; Tometsko, M.E.; Roux, E.R.; Teepe, M.C.; DuBose, R.F.; Cosman, D.; Galibert, L. A homologue of the tnf receptor and its ligand enhance t- cell growth and dendritic-cell function. Nature 1997, 390, 175-179. 72. Meednu, N.; Zhang, H.; Owen, T.; Sun, W.; Wang, V.; Cistrone, C.; Rangel-Moreno, J.; Xing, L.; Anolik, J.H. Production of rankl by memory b cells: A link between b cells and bone erosion in rheumatoid arthritis. Arthritis & rheumatology (Hoboken, N.J.) 2016, 68, 805-816. 73. Takeshita, S.; Fumoto, T.; Naoe, Y.; Ikeda, K. Age-related marrow adipogenesis is linked to increased expression of rankl. The Journal of biological chemistry 2014, 289, 16699-16710. 74. Xiong, J.; Onal, M.; Jilka, R.L.; Weinstein, R.S.; Manolagas, S.C.; O'Brien, C.A. Matrix-embedded cells control osteoclast formation. Nature medicine 2011, 17, 1235-1241. 75. Liu, C.; Walter, T.S.; Huang, P.; Zhang, S.; Zhu, X.; Wu, Y.; Wedderburn, L.R.; Tang, P.; Owens, R.J.; Stuart, D.I., et al. Structural and functional insights of rankl-rank interaction and signaling. Journal of immunology (Baltimore, Md. : 1950) 2010, 184, 6910-6919. 76. Ono, T.; Nakashima, T. Recent advances in osteoclast biology. Histochemistry and cell biology 2018, 149, 325-341. 77. Gori, F.; Hofbauer, L.C.; Dunstan, C.R.; Spelsberg, T.C.; Khosla, S.; Riggs, B.L. The expression of osteoprotegerin and rank ligand and the support of osteoclast formation by stromal-osteoblast lineage cells is developmentally regulated. Endocrinology 2000, 141, 4768-4776. 78. Udagawa, N.; Takahashi, N.; Yasuda, H.; Mizuno, A.; Itoh, K.; Ueno, Y.; Shinki, T.; Gillespie, M.T.; Martin, T.J.; Higashio, K., et al. Osteoprotegerin produced by osteoblasts is an important regulator in osteoclast development and function. Endocrinology 2000, 141, 3478-3484. 79. Kramer, I.; Halleux, C.; Keller, H.; Pegurri, M.; Gooi, J.H.; Weber, P.B.; Feng, J.Q.; Bonewald, L.F.; Kneissel, M. Osteocyte wnt/beta-catenin signaling is required for normal bone homeostasis. Molecular and cellular biology 2010, 30, 3071-3085. 80. Yasuda, H.; Shima, N.; Nakagawa, N.; Mochizuki, S.I.; Yano, K.; Fujise, N.; Sato, Y.; Goto, M.; Yamaguchi, K.; Kuriyama, M., et al. Identity of osteoclastogenesis inhibitory factor (ocif) and osteoprotegerin (opg): A mechanism by which opg/ocif inhibits osteoclastogenesis in vitro. Endocrinology 1998, 139, 1329-1337. 81. Collin-Osdoby, P.; Rothe, L.; Anderson, F.; Nelson, M.; Maloney, W.; Osdoby, P. Receptor activator of nf-kappa b and osteoprotegerin expression by human microvascular endothelial cells, regulation by inflammatory cytokines, and role in human osteoclastogenesis. The Journal of biological chemistry 2001, 276, 20659-20672. 82. Malyankar, U.M.; Scatena, M.; Suchland, K.L.; Yun, T.J.; Clark, E.A.; Giachelli, C.M. Osteoprotegerin

132

is an alpha vbeta 3-induced, nf-kappa b-dependent survival factor for endothelial cells. The Journal of biological chemistry 2000, 275, 20959-20962. 83. Yun, T.J.; Chaudhary, P.M.; Shu, G.L.; Frazer, J.K.; Ewings, M.K.; Schwartz, S.M.; Pascual, V.; Hood, L.E.; Clark, E.A. Opg/fdcr-1, a tnf receptor family member, is expressed in lymphoid cells and is upregulated by ligating cd40. Journal of immunology (Baltimore, Md. : 1950) 1998, 161, 6113-6121. 84. Walsh, M.C.; Choi, Y. Biology of the rankl-rank-opg system in immunity, bone, and beyond. Frontiers in immunology 2014, 5, 511. 85. Boyce, B.F.; Xing, L. Functions of rankl/rank/opg in bone modeling and remodeling. Archives of biochemistry and biophysics 2008, 473, 139-146. 86. Dougall, W.C.; Glaccum, M.; Charrier, K.; Rohrbach, K.; Brasel, K.; De Smedt, T.; Daro, E.; Smith, J.; Tometsko, M.E.; Maliszewski, C.R., et al. Rank is essential for osteoclast and lymph node development. Genes & development 1999, 13, 2412-2424. 87. 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., et al. Opgl is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 1999, 397, 315-323. 88. Li, J.; Sarosi, I.; Yan, X.Q.; Morony, S.; Capparelli, C.; Tan, H.L.; McCabe, S.; Elliott, R.; Scully, S.; Van, G., et al. Rank is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proceedings of the National Academy of Sciences of the United States of America 2000, 97, 1566-1571. 89. Min, H.; Morony, S.; Sarosi, I.; Dunstan, C.R.; Capparelli, C.; Scully, S.; Van, G.; Kaufman, S.; Kostenuik, P.J.; Lacey, D.L., et al. Osteoprotegerin reverses osteoporosis by inhibiting endosteal osteoclasts and prevents vascular calcification by blocking a process resembling osteoclastogenesis. The Journal of experimental medicine 2000, 192, 463-474. 90. 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., et al. Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 1997, 89, 309-319. 91. Bucay, N.; Sarosi, I.; Dunstan, C.R.; Morony, S.; Tarpley, J.; Capparelli, C.; Scully, S.; Tan, H.L.; Xu, W.; Lacey, D.L., et al. Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes & development 1998, 12, 1260-1268. 92. Mizuno, A.; Amizuka, N.; Irie, K.; Murakami, A.; Fujise, N.; Kanno, T.; Sato, Y.; Nakagawa, N.; Yasuda, H.; Mochizuki, S., et al. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochemical and biophysical research communications 1998, 247, 610-615. 93. Wright, H.L.; McCarthy, H.S.; Middleton, J.; Marshall, M.J. Rank, rankl and osteoprotegerin in bone biology and disease. Current reviews in musculoskeletal medicine 2009, 2, 56-64. 94. Lomaga, M.A.; Yeh, W.C.; Sarosi, I.; Duncan, G.S.; Furlonger, C.; Ho, A.; Morony, S.; Capparelli, C.; Van, G.; Kaufman, S., et al. Traf6 deficiency results in osteopetrosis and defective interleukin-1, cd40, and lps signaling. Genes & development 1999, 13, 1015-1024. 95. Naito, A.; Azuma, S.; Tanaka, S.; Miyazaki, T.; Takaki, S.; Takatsu, K.; Nakao, K.; Nakamura, K.; Katsuki, M.; Yamamoto, T., et al. Severe osteopetrosis, defective interleukin-1 signalling and lymph node organogenesis in traf6-deficient mice. Genes to cells : devoted to molecular & cellular mechanisms 1999, 4, 353-362. 96. Mizukami, J.; Takaesu, G.; Akatsuka, H.; Sakurai, H.; Ninomiya-Tsuji, J.; Matsumoto, K.; Sakurai, N. Receptor activator of nf-kappab ligand (rankl) activates tak1 mitogen-activated protein kinase kinase kinase through a signaling complex containing rank, tab2, and traf6. Molecular and cellular biology 2002,

133

22, 992-1000. 97. Xu, D.; Wang, S.; Liu, W.; Liu, J.; Feng, X. A novel receptor activator of nf-kappab (rank) cytoplasmic motif plays an essential role in osteoclastogenesis by committing macrophages to the osteoclast lineage. The Journal of biological chemistry 2006, 281, 4678-4690. 98. Jules, J.; Shi, Z.; Liu, J.; Xu, D.; Wang, S.; Feng, X. Receptor activator of nf-{kappa}b (rank) cytoplasmic ivvy535-538 motif plays an essential role in tumor necrosis factor-{alpha} (tnf)-mediated osteoclastogenesis. The Journal of biological chemistry 2010, 285, 37427-37435. 99. Jules, J.; Zhang, P.; Ashley, J.W.; Wei, S.; Shi, Z.; Liu, J.; Michalek, S.M.; Feng, X. Molecular basis of requirement of receptor activator of nuclear factor kappab signaling for interleukin 1-mediated osteoclastogenesis. The Journal of biological chemistry 2012, 287, 15728-15738. 100. Ghosh, S.; Karin, M. Missing pieces in the nf-kappab puzzle. Cell 2002, 109 Suppl, S81-96. 101. May, M.J.; Ghosh, S. Rel/nf-kappa b and i kappa b proteins: An overview. Seminars in cancer biology 1997, 8, 63-73. 102. Novack, D.V. Role of nf-kappab in the skeleton. Cell research 2011, 21, 169-182. 103. Senftleben, U.; Cao, Y.; Xiao, G.; Greten, F.R.; Krahn, G.; Bonizzi, G.; Chen, Y.; Hu, Y.; Fong, A.; Sun, S.C., et al. Activation by ikkalpha of a second, evolutionary conserved, nf-kappa b signaling pathway. Science (New York, N.Y.) 2001, 293, 1495-1499. 104. Yamashita, T.; Yao, Z.; Li, F.; Zhang, Q.; Badell, I.R.; Schwarz, E.M.; Takeshita, S.; Wagner, E.F.; Noda, M.; Matsuo, K., et al. Nf-kappab p50 and p52 regulate receptor activator of nf-kappab ligand (rankl) and tumor necrosis factor-induced osteoclast precursor differentiation by activating c-fos and nfatc1. The Journal of biological chemistry 2007, 282, 18245-18253. 105. Sumiya, E.; Negishi-Koga, T.; Nagai, Y.; Suematsu, A.; Suda, T.; Shinohara, M.; Sato, K.; Sanjo, H.; Akira, S.; Takayanagi, H. Phosphoproteomic analysis of kinase-deficient mice reveals multiple tak1 targets in osteoclast differentiation. Biochemical and biophysical research communications 2015, 463, 1284-1290. 106. Ruocco, M.G.; Maeda, S.; Park, J.M.; Lawrence, T.; Hsu, L.C.; Cao, Y.; Schett, G.; Wagner, E.F.; Karin, M. I{kappa}b kinase (ikk){beta}, but not ikk{alpha}, is a critical mediator of osteoclast survival and is required for inflammation-induced bone loss. The Journal of experimental medicine 2005, 201, 1677- 1687. 107. Franzoso, G.; Carlson, L.; Xing, L.; Poljak, L.; Shores, E.W.; Brown, K.D.; Leonardi, A.; Tran, T.; Boyce, B.F.; Siebenlist, U. Requirement for nf-kappab in osteoclast and b-cell development. Genes & development 1997, 11, 3482-3496. 108. Grigoriadis, A.E.; Wang, Z.Q.; Cecchini, M.G.; Hofstetter, W.; Felix, R.; Fleisch, H.A.; Wagner, E.F. C- fos: A key regulator of osteoclast-macrophage lineage determination and bone remodeling. Science (New York, N.Y.) 1994, 266, 443-448. 109. Lin, J.; Lee, D.; Choi, Y.; Lee, S.Y. The scaffold protein rack1 mediates the rankl-dependent activation of p38 mapk in osteoclast precursors. Science signaling 2015, 8, ra54. 110. Bohm, C.; Hayer, S.; Kilian, A.; Zaiss, M.M.; Finger, S.; Hess, A.; Engelke, K.; Kollias, G.; Kronke, G.; Zwerina, J., et al. The alpha-isoform of p38 mapk specifically regulates arthritic bone loss. Journal of immunology (Baltimore, Md. : 1950) 2009, 183, 5938-5947. 111. Yamamoto, A.; Miyazaki, T.; Kadono, Y.; Takayanagi, H.; Miura, T.; Nishina, H.; Katada, T.; Wakabayashi, K.; Oda, H.; Nakamura, K., et al. Possible involvement of ikappab kinase 2 and mkk7 in osteoclastogenesis induced by receptor activator of nuclear factor kappab ligand. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2002, 17,

134

612-621. 112. David, J.P.; Sabapathy, K.; Hoffmann, O.; Idarraga, M.H.; Wagner, E.F. Jnk1 modulates osteoclastogenesis through both c-jun phosphorylation-dependent and -independent mechanisms. Journal of cell science 2002, 115, 4317-4325. 113. Ikeda, F.; Nishimura, R.; Matsubara, T.; Tanaka, S.; Inoue, J.; Reddy, S.V.; Hata, K.; Yamashita, K.; Hiraga, T.; Watanabe, T., et al. Critical roles of c-jun signaling in regulation of nfat family and rankl- regulated osteoclast differentiation. The Journal of clinical investigation 2004, 114, 475-484. 114. Kenner, L.; Hoebertz, A.; Beil, F.T.; Keon, N.; Karreth, F.; Eferl, R.; Scheuch, H.; Szremska, A.; Amling, M.; Schorpp-Kistner, M., et al. Mice lacking junb are osteopenic due to cell-autonomous osteoblast and osteoclast defects. The Journal of cell biology 2004, 164, 613-623. 115. Asagiri, M.; Sato, K.; Usami, T.; Ochi, S.; Nishina, H.; Yoshida, H.; Morita, I.; Wagner, E.F.; Mak, T.W.; Serfling, E., et al. Autoamplification of nfatc1 expression determines its essential role in bone homeostasis. The Journal of experimental medicine 2005, 202, 1261-1269. 116. Cao, H.; Yu, S.; Yao, Z.; Galson, D.L.; Jiang, Y.; Zhang, X.; Fan, J.; Lu, B.; Guan, Y.; Luo, M., et al. Activating transcription factor 4 regulates osteoclast differentiation in mice. The Journal of clinical investigation 2010, 120, 2755-2766. 117. Maruyama, K.; Fukasaka, M.; Vandenbon, A.; Saitoh, T.; Kawasaki, T.; Kondo, T.; Yokoyama, K.K.; Kidoya, H.; Takakura, N.; Standley, D., et al. The transcription factor jdp2 controls bone homeostasis and antibacterial immunity by regulating osteoclast and neutrophil differentiation. Immunity 2012, 37, 1024-1036. 118. Matsuo, K.; Galson, D.L.; Zhao, C.; Peng, L.; Laplace, C.; Wang, K.Z.; Bachler, M.A.; Amano, H.; Aburatani, H.; Ishikawa, H., et al. Nuclear factor of activated t-cells (nfat) rescues osteoclastogenesis in precursors lacking c-fos. The Journal of biological chemistry 2004, 279, 26475-26480. 119. Aliprantis, A.O.; Ueki, Y.; Sulyanto, R.; Park, A.; Sigrist, K.S.; Sharma, S.M.; Ostrowski, M.C.; Olsen, B.R.; Glimcher, L.H. Nfatc1 in mice represses osteoprotegerin during osteoclastogenesis and dissociates systemic osteopenia from inflammation in cherubism. The Journal of clinical investigation 2008, 118, 3775-3789. 120. Takayanagi, H.; Kim, S.; Koga, T.; Nishina, H.; Isshiki, M.; Yoshida, H.; Saiura, A.; Isobe, M.; Yokochi, T.; Inoue, J., et al. Induction and activation of the transcription factor nfatc1 (nfat2) integrate rankl signaling in terminal differentiation of osteoclasts. Developmental cell 2002, 3, 889-901. 121. Kim, K.; Lee, S.H.; Ha Kim, J.; Choi, Y.; Kim, N. Nfatc1 induces osteoclast fusion via up-regulation of atp6v0d2 and the dendritic cell-specific transmembrane protein (dc-stamp). Molecular endocrinology (Baltimore, Md.) 2008, 22, 176-185. 122. Pang, M.; Rodriguez-Gonzalez, M.; Hernandez, M.; Recinos, C.C.; Seldeen, K.L.; Troen, B.R. Ap-1 and mitf interact with nfatc1 to stimulate cathepsin k promoter activity in osteoclast precursors. Journal of cellular biochemistry 2019, 120, 12382-12392. 123. Kim, Y.; Sato, K.; Asagiri, M.; Morita, I.; Soma, K.; Takayanagi, H. Contribution of nuclear factor of activated t cells c1 to the transcriptional control of immunoreceptor osteoclast-associated receptor but not triggering receptor expressed by myeloid cells-2 during osteoclastogenesis. The Journal of biological chemistry 2005, 280, 32905-32913. 124. Crotti, T.N.; Flannery, M.; Walsh, N.C.; Fleming, J.D.; Goldring, S.R.; McHugh, K.P. Nfatc1 regulation of the human beta3 integrin promoter in osteoclast differentiation. Gene 2006, 372, 92-102. 125. Kim, J.H.; Kim, N. Regulation of nfatc1 in osteoclast differentiation. Journal of bone metabolism 2014, 21, 233-241.

135

126. Kim, J.H.; Kim, N. Signaling pathways in osteoclast differentiation. Chonnam medical journal 2016, 52, 12-17. 127. Park, J.H.; Lee, N.K.; Lee, S.Y. Current understanding of rank signaling in osteoclast differentiation and maturation. Molecules and cells 2017, 40, 706-713. 128. Koga, T.; Inui, M.; Inoue, K.; Kim, S.; Suematsu, A.; Kobayashi, E.; Iwata, T.; Ohnishi, H.; Matozaki, T.; Kodama, T., et al. Costimulatory signals mediated by the itam motif cooperate with rankl for bone homeostasis. Nature 2004, 428, 758-763. 129. Shinohara, M.; Koga, T.; Okamoto, K.; Sakaguchi, S.; Arai, K.; Yasuda, H.; Takai, T.; Kodama, T.; Morio, T.; Geha, R.S., et al. Tyrosine kinases btk and tec regulate osteoclast differentiation by linking rank and itam signals. Cell 2008, 132, 794-806. 130. Yang, Y.M.; Kim, M.S.; Son, A.; Hong, J.H.; Kim, K.H.; Seo, J.T.; Lee, S.I.; Shin, D.M. Alteration of rankl-induced osteoclastogenesis in primary cultured osteoclasts from serca2+/- mice. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2009, 24, 1763-1769. 131. Vignery, A. Macrophage fusion: The making of osteoclasts and giant cells. The Journal of experimental medicine 2005, 202, 337-340. 132. Irie, A.; Yamamoto, K.; Miki, Y.; Murakami, M. Phosphatidylethanolamine dynamics are required for osteoclast fusion. Scientific reports 2017, 7, 46715. 133. Chiu, Y.H.; Ritchlin, C.T. Dc-stamp: A key regulator in osteoclast differentiation. Journal of cellular physiology 2016, 231, 2402-2407. 134. Kukita, T.; Wada, N.; Kukita, A.; Kakimoto, T.; Sandra, F.; Toh, K.; Nagata, K.; Iijima, T.; Horiuchi, M.; Matsusaki, H., et al. Rankl-induced dc-stamp is essential for osteoclastogenesis. The Journal of experimental medicine 2004, 200, 941-946. 135. Yagi, M.; Miyamoto, T.; Sawatani, Y.; Iwamoto, K.; Hosogane, N.; Fujita, N.; Morita, K.; Ninomiya, K.; Suzuki, T.; Miyamoto, K., et al. Dc-stamp is essential for cell-cell fusion in osteoclasts and foreign body giant cells. The Journal of experimental medicine 2005, 202, 345-351. 136. Mensah, K.A.; Ritchlin, C.T.; Schwarz, E.M. Rankl induces heterogeneous dc-stamp(lo) and dc-stamp(hi) osteoclast precursors of which the dc-stamp(lo) precursors are the master fusogens. Journal of cellular physiology 2010, 223, 76-83. 137. Yang, M.; Birnbaum, M.J.; MacKay, C.A.; Mason-Savas, A.; Thompson, B.; Odgren, P.R. Osteoclast stimulatory transmembrane protein (oc-stamp), a novel protein induced by rankl that promotes osteoclast differentiation. Journal of cellular physiology 2008, 215, 497-505. 138. Miyamoto, H.; Suzuki, T.; Miyauchi, Y.; Iwasaki, R.; Kobayashi, T.; Sato, Y.; Miyamoto, K.; Hoshi, H.; Hashimoto, K.; Yoshida, S., et al. Osteoclast stimulatory transmembrane protein and dendritic cell- specific transmembrane protein cooperatively modulate cell-cell fusion to form osteoclasts and foreign body giant cells. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2012, 27, 1289-1297. 139. Witwicka, H.; Hwang, S.Y.; Reyes-Gutierrez, P.; Jia, H.; Odgren, P.E.; Donahue, L.R.; Birnbaum, M.J.; Odgren, P.R. Studies of oc-stamp in osteoclast fusion: A new knockout mouse model, rescue of cell fusion, and transmembrane topology. PloS one 2015, 10, e0128275. 140. Lee, S.H.; Rho, J.; Jeong, D.; Sul, J.Y.; Kim, T.; Kim, N.; Kang, J.S.; Miyamoto, T.; Suda, T.; Lee, S.K., et al. V-atpase v0 subunit d2-deficient mice exhibit impaired osteoclast fusion and increased bone formation. Nature medicine 2006, 12, 1403-1409. 141. Linder, S.; Wiesner, C. Tools of the trade: Podosomes as multipurpose organelles of monocytic cells.

136

Cellular and molecular life sciences : CMLS 2015, 72, 121-135. 142. Hu, S.; Planus, E.; Georgess, D.; Place, C.; Wang, X.; Albiges-Rizo, C.; Jurdic, P.; Geminard, J.C. Podosome rings generate forces that drive saltatory osteoclast migration. Molecular biology of the cell 2011, 22, 3120-3126. 143. Luxenburg, C.; Geblinger, D.; Klein, E.; Anderson, K.; Hanein, D.; Geiger, B.; Addadi, L. The architecture of the adhesive apparatus of cultured osteoclasts: From podosome formation to sealing zone assembly. PloS one 2007, 2, e179. 144. Chabadel, A.; Banon-Rodriguez, I.; Cluet, D.; Rudkin, B.B.; Wehrle-Haller, B.; Genot, E.; Jurdic, P.; Anton, I.M.; Saltel, F. Cd44 and beta3 integrin organize two functionally distinct actin-based domains in osteoclasts. Molecular biology of the cell 2007, 18, 4899-4910. 145. Linder, S.; Kopp, P. Podosomes at a glance. Journal of cell science 2005, 118, 2079-2082. 146. Soysa, N.S.; Alles, N. Osteoclast function and bone-resorbing activity: An overview. Biochemical and biophysical research communications 2016, 476, 115-120. 147. Jurdic, P.; Saltel, F.; Chabadel, A.; Destaing, O. Podosome and sealing zone: Specificity of the osteoclast model. European journal of cell biology 2006, 85, 195-202. 148. Destaing, O.; Saltel, F.; Geminard, J.C.; Jurdic, P.; Bard, F. Podosomes display actin turnover and dynamic self-organization in osteoclasts expressing actin-green fluorescent protein. Molecular biology of the cell 2003, 14, 407-416. 149. McMichael, B.K.; Cheney, R.E.; Lee, B.S. Myosin x regulates sealing zone patterning in osteoclasts through linkage of podosomes and microtubules. The Journal of biological chemistry 2010, 285, 9506- 9515. 150. Saltel, F.; Destaing, O.; Bard, F.; Eichert, D.; Jurdic, P. Apatite-mediated actin dynamics in resorbing osteoclasts. Molecular biology of the cell 2004, 15, 5231-5241. 151. Wennerberg, K.; Der, C.J. Rho-family gtpases: It's not only rac and rho (and i like it). Journal of cell science 2004, 117, 1301-1312. 152. Lawson, C.D.; Burridge, K. The on-off relationship of rho and rac during integrin-mediated adhesion and cell migration. Small GTPases 2014, 5, e27958. 153. Sakai, H.; Chen, Y.; Itokawa, T.; Yu, K.P.; Zhu, M.L.; Insogna, K. Activated c-fms recruits vav and rac during csf-1-induced cytoskeletal remodeling and spreading in osteoclasts. Bone 2006, 39, 1290-1301. 154. Izawa, T.; Zou, W.; Chappel, J.C.; Ashley, J.W.; Feng, X.; Teitelbaum, S.L. C-src links a rank/alphavbeta3 integrin complex to the osteoclast cytoskeleton. Molecular and cellular biology 2012, 32, 2943-2953. 155. Touaitahuata, H.; Blangy, A.; Vives, V. Modulation of osteoclast differentiation and bone resorption by rho gtpases. Small GTPases 2014, 5, e28119. 156. Chellaiah, M.A.; Soga, N.; Swanson, S.; McAllister, S.; Alvarez, U.; Wang, D.; Dowdy, S.F.; Hruska, K.A. Rho-a is critical for osteoclast podosome organization, motility, and bone resorption. The Journal of biological chemistry 2000, 275, 11993-12002. 157. Ory, S.; Munari-Silem, Y.; Fort, P.; Jurdic, P. Rho and rac exert antagonistic functions on spreading of macrophage-derived multinucleated cells and are not required for actin fiber formation. Journal of cell science 2000, 113 ( Pt 7), 1177-1188. 158. Ory, S.; Brazier, H.; Pawlak, G.; Blangy, A. Rho gtpases in osteoclasts: Orchestrators of podosome arrangement. European journal of cell biology 2008, 87, 469-477. 159. Destaing, O.; Saltel, F.; Gilquin, B.; Chabadel, A.; Khochbin, S.; Ory, S.; Jurdic, P. A novel rho-mdia2- hdac6 pathway controls podosome patterning through microtubule acetylation in osteoclasts. Journal of

137

cell science 2005, 118, 2901-2911. 160. McMichael, B.K.; Scherer, K.F.; Franklin, N.C.; Lee, B.S. The rhogap activity of myosin ixb is critical for osteoclast podosome patterning, motility, and resorptive capacity. PloS one 2014, 9, e87402. 161. Croke, M.; Ross, F.P.; Korhonen, M.; Williams, D.A.; Zou, W.; Teitelbaum, S.L. Rac deletion in osteoclasts causes severe osteopetrosis. Journal of cell science 2011, 124, 3811-3821. 162. Faccio, R.; Teitelbaum, S.L.; Fujikawa, K.; Chappel, J.; Zallone, A.; Tybulewicz, V.L.; Ross, F.P.; Swat, W. Vav3 regulates osteoclast function and bone mass. Nature medicine 2005, 11, 284-290. 163. Takegahara, N.; Kang, S.; Nojima, S.; Takamatsu, H.; Okuno, T.; Kikutani, H.; Toyofuku, T.; Kumanogoh, A. Integral roles of a guanine nucleotide exchange factor, farp2, in osteoclast podosome rearrangements. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2010, 24, 4782-4792. 164. Vives, V.; Laurin, M.; Cres, G.; Larrousse, P.; Morichaud, Z.; Noel, D.; Cote, J.F.; Blangy, A. The rac1 exchange factor dock5 is essential for bone resorption by osteoclasts. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2011, 26, 1099- 1110. 165. Ito, Y.; Teitelbaum, S.L.; Zou, W.; Zheng, Y.; Johnson, J.F.; Chappel, J.; Ross, F.P.; Zhao, H. Cdc42 regulates bone modeling and remodeling in mice by modulating rankl/m-csf signaling and osteoclast polarization. The Journal of clinical investigation 2010, 120, 1981-1993. 166. Chellaiah, M.A. Regulation of actin ring formation by rho gtpases in osteoclasts. The Journal of biological chemistry 2005, 280, 32930-32943. 167. Gay, C.V.; Mueller, W.J. Carbonic anhydrase and osteoclasts: Localization by labeled inhibitor autoradiography. Science (New York, N.Y.) 1974, 183, 432-434. 168. Qin, A.; Cheng, T.S.; Pavlos, N.J.; Lin, Z.; Dai, K.R.; Zheng, M.H. V-atpases in osteoclasts: Structure, function and potential inhibitors of bone resorption. The international journal of biochemistry & cell biology 2012, 44, 1422-1435. 169. Kornak, U.; Kasper, D.; Bosl, M.R.; Kaiser, E.; Schweizer, M.; Schulz, A.; Friedrich, W.; Delling, G.; Jentsch, T.J. Loss of the clc-7 chloride channel leads to osteopetrosis in mice and man. Cell 2001, 104, 205-215. 170. Wu, J.; Glimcher, L.H.; Aliprantis, A.O. Hco3-/cl- anion exchanger slc4a2 is required for proper osteoclast differentiation and function. Proceedings of the National Academy of Sciences of the United States of America 2008, 105, 16934-16939. 171. Lotinun, S.; Kiviranta, R.; Matsubara, T.; Alzate, J.A.; Neff, L.; Luth, A.; Koskivirta, I.; Kleuser, B.; Vacher, J.; Vuorio, E., et al. Osteoclast-specific cathepsin k deletion stimulates s1p-dependent bone formation. The Journal of clinical investigation 2013, 123, 666-681. 172. Gu, J.H.; Tong, X.S.; Chen, G.H.; Liu, X.Z.; Bian, J.C.; Yuan, Y.; Liu, Z.P. Regulation of matrix metalloproteinase-9 protein expression by 1alpha, 25-(oh)(2)d(3) during osteoclast differentiation. Journal of veterinary science 2014, 15, 133-140. 173. Stenbeck, G. Formation and function of the ruffled border in osteoclasts. Seminars in cell & developmental biology 2002, 13, 285-292. 174. Stenbeck, G.; Coxon, F.P. Role of vesicular trafficking in skeletal dynamics. Current opinion in pharmacology 2014, 16, 7-14. 175. Zhao, H.; Ito, Y.; Chappel, J.; Andrews, N.W.; Teitelbaum, S.L.; Ross, F.P. Synaptotagmin vii regulates bone remodeling by modulating osteoclast and osteoblast secretion. Developmental cell 2008, 14, 914- 925.

138

176. Zhao, H. Membrane trafficking in osteoblasts and osteoclasts: New avenues for understanding and treating skeletal diseases. Traffic (Copenhagen, Denmark) 2012, 13, 1307-1314. 177. Takito, J.; Inoue, S.; Nakamura, M. The sealing zone in osteoclasts: A self-organized structure on the bone. International journal of molecular sciences 2018, 19. 178. Marino, S.; Logan, J.G.; Mellis, D.; Capulli, M. Generation and culture of osteoclasts. BoneKEy reports 2014, 3, 570. 179. Collin-Osdoby, P.; Osdoby, P. Isolation and culture of primary chicken osteoclasts. Methods in molecular biology (Clifton, N.J.) 2012, 816, 119-143. 180. Coxon, F.P.; Rogers, M.J.; Crockett, J.C. Isolation and purification of rabbit osteoclasts. Methods in molecular biology (Clifton, N.J.) 2012, 816, 145-158. 181. Itzstein, C.; van 't Hof, R.J. Osteoclast formation in mouse co-cultures. Methods in molecular biology (Clifton, N.J.) 2012, 816, 177-186. 182. Orriss, I.R.; Arnett, T.R. Rodent osteoclast cultures. Methods in molecular biology (Clifton, N.J.) 2012, 816, 103-117. 183. Henriksen, K.; Karsdal, M.A.; Taylor, A.; Tosh, D.; Coxon, F.P. Generation of human osteoclasts from peripheral blood. Methods in molecular biology (Clifton, N.J.) 2012, 816, 159-175. 184. Collin-Osdoby, P.; Osdoby, P. Rankl-mediated osteoclast formation from murine raw 264.7 cells. Methods in molecular biology (Clifton, N.J.) 2012, 816, 187-202. 185. Islam, S.; Hassan, F.; Tumurkhuu, G.; Dagvadorj, J.; Koide, N.; Naiki, Y.; Yoshida, T.; Yokochi, T. Receptor activator of nuclear factor-kappa b ligand induces osteoclast formation in raw 264.7 macrophage cells via augmented production of macrophage-colony-stimulating factor. Microbiology and immunology 2008, 52, 585-590. 186. Taciak, B.; Bialasek, M.; Braniewska, A.; Sas, Z.; Sawicka, P.; Kiraga, L.; Rygiel, T.; Krol, M. Evaluation of phenotypic and functional stability of raw 264.7 cell line through serial passages. PloS one 2018, 13, e0198943. 187. Ng, A.Y.; Tu, C.; Shen, S.; Xu, D.; Oursler, M.J.; Qu, J.; Yang, S. Comparative characterization of osteoclasts derived from murine bone marrow macrophages and raw 264.7 cells using quantitative proteomics. JBMR plus 2018, 2, 328-340. 188. Iismaa, S.E.; Mearns, B.M.; Lorand, L.; Graham, R.M. Transglutaminases and disease: Lessons from genetically engineered mouse models and inherited disorders. Physiological reviews 2009, 89, 991-1023. 189. Grenard, P.; Bates, M.K.; Aeschlimann, D. Evolution of transglutaminase genes: Identification of a transglutaminase gene cluster on human chromosome 15q15. Structure of the gene encoding transglutaminase x and a novel gene family member, transglutaminase z. The Journal of biological chemistry 2001, 276, 33066-33078. 190. Eckert, R.L.; Kaartinen, M.T.; Nurminskaya, M.; Belkin, A.M.; Colak, G.; Johnson, G.V.; Mehta, K. Transglutaminase regulation of cell function. Physiological reviews 2014, 94, 383-417. 191. Lee, C.S.; Park, H.H. Structural aspects of transglutaminase 2: Functional, structural, and regulatory diversity. Apoptosis : an international journal on programmed cell death 2017, 22, 1057-1068. 192. Katt, W.P.; Antonyak, M.A.; Cerione, R.A. Opening up about tissue transglutaminase: When conformation matters more than enzymatic activity. Med one 2018, 3. 193. Zhang, J.; Antonyak, M.A.; Singh, G.; Cerione, R.A. A mechanism for the upregulation of egf receptor levels in glioblastomas. Cell reports 2013, 3, 2008-2020. 194. Kerr, C.; Szmacinski, H.; Fisher, M.L.; Nance, B.; Lakowicz, J.R.; Akbar, A.; Keillor, J.W.; Lok Wong, T.; Godoy-Ruiz, R.; Toth, E.A., et al. Transamidase site-targeted agents alter the conformation of the

139

transglutaminase cancer stem cell survival protein to reduce gtp binding activity and cancer stem cell survival. Oncogene 2017, 36, 2981-2990. 195. Caron, N.S.; Munsie, L.N.; Keillor, J.W.; Truant, R. Using flim-fret to measure conformational changes of transglutaminase type 2 in live cells. PloS one 2012, 7, e44159. 196. Lorand, L.; Graham, R.M. Transglutaminases: Crosslinking enzymes with pleiotropic functions. Nature reviews. Molecular cell biology 2003, 4, 140-156. 197. Moldogazieva, N.T.; Mokhosoev, I.M.; Feldman, N.B.; Lutsenko, S.V. Ros and rns signalling: Adaptive redox switches through oxidative/nitrosative protein modifications. Free radical research 2018, 52, 507- 543. 198. Ray, P.D.; Huang, B.W.; Tsuji, Y. Reactive oxygen species (ros) homeostasis and redox regulation in cellular signaling. Cellular signalling 2012, 24, 981-990. 199. Poljsak, B.; Suput, D.; Milisav, I. Achieving the balance between ros and antioxidants: When to use the synthetic antioxidants. Oxidative medicine and cellular longevity 2013, 2013, 956792. 200. Birben, E.; Sahiner, U.M.; Sackesen, C.; Erzurum, S.; Kalayci, O. Oxidative stress and antioxidant defense. The World Allergy Organization journal 2012, 5, 9-19. 201. Kiraly, R.; Demeny, M.; Fesus, L. Protein transamidation by transglutaminase 2 in cells: A disputed ca2+- dependent action of a multifunctional protein. The FEBS journal 2011, 278, 4717-4739. 202. Stamnaes, J.; Pinkas, D.M.; Fleckenstein, B.; Khosla, C.; Sollid, L.M. Redox regulation of transglutaminase 2 activity. The Journal of biological chemistry 2010, 285, 25402-25409. 203. Siegel, M.; Strnad, P.; Watts, R.E.; Choi, K.; Jabri, B.; Omary, M.B.; Khosla, C. Extracellular transglutaminase 2 is catalytically inactive, but is transiently activated upon tissue injury. PloS one 2008, 3, e1861. 204. Jin, X.; Stamnaes, J.; Klock, C.; DiRaimondo, T.R.; Sollid, L.M.; Khosla, C. Activation of extracellular transglutaminase 2 by thioredoxin. The Journal of biological chemistry 2011, 286, 37866-37873. 205. Leon, L.; Jeannin, J.F.; Bettaieb, A. Post-translational modifications induced by nitric oxide (no): Implication in cancer cells apoptosis. Nitric oxide : biology and chemistry 2008, 19, 77-83. 206. Lai, T.S.; Hausladen, A.; Slaughter, T.F.; Eu, J.P.; Stamler, J.S.; Greenberg, C.S. Calcium regulates s- nitrosylation, denitrosylation, and activity of tissue transglutaminase. Biochemistry 2001, 40, 4904-4910. 207. Melino, G.; Bernassola, F.; Knight, R.A.; Corasaniti, M.T.; Nistico, G.; Finazzi-Agro, A. S-nitrosylation regulates apoptosis. Nature 1997, 388, 432-433. 208. Santhanam, L.; Tuday, E.C.; Webb, A.K.; Dowzicky, P.; Kim, J.H.; Oh, Y.J.; Sikka, G.; Kuo, M.; Halushka, M.K.; Macgregor, A.M., et al. Decreased s-nitrosylation of tissue transglutaminase contributes to age-related increases in vascular stiffness. Circulation research 2010, 107, 117-125. 209. Telci, D.; Collighan, R.J.; Basaga, H.; Griffin, M. Increased tg2 expression can result in induction of transforming growth factor beta1, causing increased synthesis and deposition of matrix proteins, which can be regulated by nitric oxide. The Journal of biological chemistry 2009, 284, 29547-29558. 210. Begg, G.E.; Carrington, L.; Stokes, P.H.; Matthews, J.M.; Wouters, M.A.; Husain, A.; Lorand, L.; Iismaa, S.E.; Graham, R.M. Mechanism of allosteric regulation of transglutaminase 2 by gtp. Proceedings of the National Academy of Sciences of the United States of America 2006, 103, 19683-19688. 211. Sugimura, Y.; Hosono, M.; Kitamura, M.; Tsuda, T.; Yamanishi, K.; Maki, M.; Hitomi, K. Identification of preferred substrate sequences for transglutaminase 1--development of a novel peptide that can efficiently detect cross-linking enzyme activity in the skin. The FEBS journal 2008, 275, 5667-5677. 212. Sugimura, Y.; Hosono, M.; Wada, F.; Yoshimura, T.; Maki, M.; Hitomi, K. Screening for the preferred substrate sequence of transglutaminase using a phage-displayed peptide library: Identification of peptide

140

substrates for tgase 2 and factor xiiia. The Journal of biological chemistry 2006, 281, 17699-17706. 213. Hitomi, K.; Kitamura, M.; Sugimura, Y. Preferred substrate sequences for transglutaminase 2: Screening using a phage-displayed peptide library. Amino acids 2009, 36, 619-624. 214. Yamane, A.; Fukui, M.; Sugimura, Y.; Itoh, M.; Alea, M.P.; Thomas, V.; El Alaoui, S.; Akiyama, M.; Hitomi, K. Identification of a preferred substrate peptide for transglutaminase 3 and detection of in situ activity in skin and hair follicles. The FEBS journal 2010, 277, 3564-3574. 215. Fukui, M.; Kuramoto, K.; Yamasaki, R.; Shimizu, Y.; Itoh, M.; Kawamoto, T.; Hitomi, K. Identification of a highly reactive substrate peptide for transglutaminase 6 and its use in detecting transglutaminase activity in the skin epidermis. The FEBS journal 2013, 280, 1420-1429. 216. Kuramoto, K.; Yamasaki, R.; Shimizu, Y.; Tatsukawa, H.; Hitomi, K. Phage-displayed peptide library screening for preferred human substrate peptide sequences for transglutaminase 7. Archives of biochemistry and biophysics 2013, 537, 138-143. 217. Hitomi, K.; Kitamura, M.; Alea, M.P.; Ceylan, I.; Thomas, V.; El Alaoui, S. A specific colorimetric assay for measuring transglutaminase 1 and factor xiii activities. Analytical biochemistry 2009, 394, 281-283. 218. Perez Alea, M.; Kitamura, M.; Martin, G.; Thomas, V.; Hitomi, K.; El Alaoui, S. Development of an isoenzyme-specific colorimetric assay for tissue transglutaminase 2 cross-linking activity. Analytical biochemistry 2009, 389, 150-156. 219. Itoh, M.; Kawamoto, T.; Tatsukawa, H.; Kojima, S.; Yamanishi, K.; Hitomi, K. In situ detection of active transglutaminases for keratinocyte type (tgase 1) and tissue type (tgase 2) using fluorescence-labeled highly reactive substrate peptides. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 2011, 59, 180-187. 220. Itoh, M.; Tatsukawa, H.; Eun-Seo, L.; Yamanishi, K.; Kojima, S.; Hitomi, K. Variations in both tg1 and tg2 isozyme-specific in situ activities and protein expressions during mouse embryonic development. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 2013, 61, 793-801. 221. Watanabe, K.; Tsunoda, K.; Itoh, M.; Fukui, M.; Mori, H.; Hitomi, K. Transglutaminase 2 and factor xiii catalyze distinct substrates in differentiating osteoblastic cell line: Utility of highly reactive substrate peptides. Amino acids 2013, 44, 209-214. 222. Tatsukawa, H.; Otsu, R.; Tani, Y.; Wakita, R.; Hitomi, K. Isozyme-specific comprehensive characterization of transglutaminase-crosslinked substrates in kidney fibrosis. Scientific reports 2018, 8, 7306. 223. Tatsukawa, H.; Tani, Y.; Otsu, R.; Nakagawa, H.; Hitomi, K. Global identification and analysis of isozyme-specific possible substrates crosslinked by transglutaminases using substrate peptides in mouse liver fibrosis. Scientific reports 2017, 7, 45049. 224. Katt, W.P.; Antonyak, M.A.; Cerione, R.A. The diamond anniversary of tissue transglutaminase: A protein of many talents. Drug discovery today 2018, 23, 575-591. 225. Lee, K.N.; Fesus, L.; Yancey, S.T.; Girard, J.E.; Chung, S.I. Development of selective inhibitors of transglutaminase. Phenylthiourea derivatives. The Journal of biological chemistry 1985, 260, 14689- 14694. 226. Keillor, J.W.; Apperley, K.Y. Transglutaminase inhibitors: A patent review. Expert opinion on therapeutic patents 2016, 26, 49-63. 227. Keillor, J.W.; Apperley, K.Y.; Akbar, A. Inhibitors of tissue transglutaminase. Trends in pharmacological sciences 2015, 36, 32-40. 228. Schaertl, S.; Prime, M.; Wityak, J.; Dominguez, C.; Munoz-Sanjuan, I.; Pacifici, R.E.; Courtney, S.;

141

Scheel, A.; Macdonald, D. A profiling platform for the characterization of transglutaminase 2 (tg2) inhibitors. Journal of biomolecular screening 2010, 15, 478-487. 229. Yazdani, Y.; Azari, S.; Kalhor, H.R. Expression of functional recombinant human tissue transglutaminase (tg2) using the bac-to-bac baculovirus expression system. Advanced pharmaceutical bulletin 2016, 6, 49-56. 230. Adamczyk, M.; Griffiths, R.; Dewitt, S.; Knauper, V.; Aeschlimann, D. P2x7 receptor activation regulates rapid unconventional export of transglutaminase-2. Journal of cell science 2015, 128, 4615-4628. 231. Aeschlimann, D.; Knauper, V. P2x7 receptor-mediated tg2 externalization: A link to inflammatory arthritis? Amino acids 2017, 49, 453-460. 232. Santhanam, L.; Berkowitz, D.E.; Belkin, A.M. Nitric oxide regulates non-classical secretion of tissue transglutaminase. Communicative & integrative biology 2011, 4, 584-586. 233. Zemskov, E.A.; Mikhailenko, I.; Hsia, R.C.; Zaritskaya, L.; Belkin, A.M. Unconventional secretion of tissue transglutaminase involves phospholipid-dependent delivery into recycling endosomes. PloS one 2011, 6, e19414. 234. Scarpellini, A.; Germack, R.; Lortat-Jacob, H.; Muramatsu, T.; Billett, E.; Johnson, T.; Verderio, E.A. Heparan sulfate proteoglycans are receptors for the cell-surface trafficking and biological activity of transglutaminase-2. The Journal of biological chemistry 2009, 284, 18411-18423. 235. Szondy, Z.; Korponay-Szabo, I.; Kiraly, R.; Sarang, Z.; Tsay, G.J. Transglutaminase 2 in human diseases. BioMedicine 2017, 7, 15. 236. Rauhavirta, T.; Hietikko, M.; Salmi, T.; Lindfors, K. Transglutaminase 2 and transglutaminase 2 autoantibodies in celiac disease: A review. Clinical reviews in allergy & immunology 2016. 237. Nakaoka, H.; Perez, D.M.; Baek, K.J.; Das, T.; Husain, A.; Misono, K.; Im, M.J.; Graham, R.M. Gh: A gtp-binding protein with transglutaminase activity and receptor signaling function. Science (New York, N.Y.) 1994, 264, 1593-1596. 238. Nurminskaya, M.V.; Belkin, A.M. Cellular functions of tissue transglutaminase. International review of cell and molecular biology 2012, 294, 1-97. 239. Nakano, Y.; Addison, W.N.; Kaartinen, M.T. Atp-mediated mineralization of mc3t3-e1 osteoblast cultures. Bone 2007, 41, 549-561. 240. Nakano, Y.; Forsprecher, J.; Kaartinen, M.T. Regulation of atpase activity of transglutaminase 2 by mt1- mmp: Implications for mineralization of mc3t3-e1 osteoblast cultures. Journal of cellular physiology 2010, 223, 260-269. 241. Malorni, W.; Farrace, M.G.; Matarrese, P.; Tinari, A.; Ciarlo, L.; Mousavi-Shafaei, P.; D'Eletto, M.; Di Giacomo, G.; Melino, G.; Palmieri, L., et al. The adenine nucleotide translocator 1 acts as a type 2 transglutaminase substrate: Implications for mitochondrial-dependent apoptosis. Cell death and differentiation 2009, 16, 1480-1492. 242. Mishra, S.; Murphy, L.J. Tissue transglutaminase has intrinsic kinase activity: Identification of transglutaminase 2 as an insulin-like growth factor-binding protein-3 kinase. The Journal of biological chemistry 2004, 279, 23863-23868. 243. Mishra, S.; Murphy, L.J. The p53 oncoprotein is a substrate for tissue transglutaminase kinase activity. Biochemical and biophysical research communications 2006, 339, 726-730. 244. Akimov, S.S.; Krylov, D.; Fleischman, L.F.; Belkin, A.M. Tissue transglutaminase is an integrin-binding adhesion coreceptor for fibronectin. The Journal of cell biology 2000, 148, 825-838. 245. Belkin, A.M. Extracellular tg2: Emerging functions and regulation. The FEBS journal 2011, 278, 4704- 4716.

142

246. De Laurenzi, V.; Melino, G. Gene disruption of tissue transglutaminase. Molecular and cellular biology 2001, 21, 148-155. 247. Falasca, L.; Iadevaia, V.; Ciccosanti, F.; Melino, G.; Serafino, A.; Piacentini, M. Transglutaminase type ii is a key element in the regulation of the anti-inflammatory response elicited by apoptotic cell engulfment. Journal of immunology (Baltimore, Md. : 1950) 2005, 174, 7330-7340. 248. Szondy, Z.; Sarang, Z.; Molnar, P.; Nemeth, T.; Piacentini, M.; Mastroberardino, P.G.; Falasca, L.; Aeschlimann, D.; Kovacs, J.; Kiss, I., et al. Transglutaminase 2-/- mice reveal a phagocytosis-associated crosstalk between macrophages and apoptotic cells. Proceedings of the National Academy of Sciences of the United States of America 2003, 100, 7812-7817. 249. Bernassola, F.; Federici, M.; Corazzari, M.; Terrinoni, A.; Hribal, M.L.; De Laurenzi, V.; Ranalli, M.; Massa, O.; Sesti, G.; McLean, W.H., et al. Role of transglutaminase 2 in glucose tolerance: Knockout mice studies and a putative mutation in a mody patient. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 2002, 16, 1371-1378. 250. Szondy, Z.; Mastroberardino, P.G.; Varadi, J.; Farrace, M.G.; Nagy, N.; Bak, I.; Viti, I.; Wieckowski, M.R.; Melino, G.; Rizzuto, R., et al. Tissue transglutaminase (tg2) protects cardiomyocytes against ischemia/reperfusion injury by regulating atp synthesis. Cell death and differentiation 2006, 13, 1827- 1829. 251. Olsen, K.C.; Sapinoro, R.E.; Kottmann, R.M.; Kulkarni, A.A.; Iismaa, S.E.; Johnson, G.V.; Thatcher, T.H.; Phipps, R.P.; Sime, P.J. Transglutaminase 2 and its role in pulmonary fibrosis. American journal of respiratory and critical care medicine 2011, 184, 699-707. 252. Katona, E.; Penzes, K.; Csapo, A.; Fazakas, F.; Udvardy, M.L.; Bagoly, Z.; Orosz, Z.Z.; Muszbek, L. Interaction of factor xiii subunits. Blood 2014, 123, 1757-1763. 253. Muszbek, L.; Bereczky, Z.; Bagoly, Z.; Komaromi, I.; Katona, E. Factor xiii: A coagulation factor with multiple plasmatic and cellular functions. Physiological reviews 2011, 91, 931-972. 254. Dorgalaleh, A.; Rashidpanah, J. Blood coagulation factor xiii and factor xiii deficiency. Blood reviews 2016, 30, 461-475. 255. Moaddel, M.; Farrell, D.H.; Daugherty, M.A.; Fried, M.G. Interactions of human fibrinogens with factor xiii: Roles of calcium and the gamma' peptide. Biochemistry 2000, 39, 6698-6705. 256. Mitchell, J.L.; Mutch, N.J. Let's cross-link: Diverse functions of the promiscuous cellular transglutaminase factor xiii-a. Journal of thrombosis and haemostasis : JTH 2019, 17, 19-30. 257. Beckers, C.M.L.; Simpson, K.R.; Griffin, K.J.; Brown, J.M.; Cheah, L.T.; Smith, K.A.; Vacher, J.; Cordell, P.A.; Kearney, M.T.; Grant, P.J., et al. Cre/lox studies identify resident macrophages as the major source of circulating coagulation factor xiii-a. Arteriosclerosis, thrombosis, and vascular biology 2017, 37, 1494-1502. 258. Polgar, J.; Hidasi, V.; Muszbek, L. Non-proteolytic activation of cellular protransglutaminase (placenta macrophage factor xiii). The Biochemical journal 1990, 267, 557-560. 259. Cordell, P.A.; Kile, B.T.; Standeven, K.F.; Josefsson, E.C.; Pease, R.J.; Grant, P.J. Association of coagulation factor xiii-a with golgi proteins within monocyte-macrophages: Implications for subcellular trafficking and secretion. Blood 2010, 115, 2674-2681. 260. Myneni, V.D.; Mousa, A.; Kaartinen, M.T. Factor xiii-a transglutaminase deficient mice show signs of metabolically healthy obesity on high fat diet. Scientific reports 2016, 6, 35574. 261. Koseki-Kuno, S.; Yamakawa, M.; Dickneite, G.; Ichinose, A. Factor xiii a subunit-deficient mice developed severe uterine bleeding events and subsequent spontaneous miscarriages. Blood 2003, 102, 4410-4412.

143

262. Nahrendorf, M.; Hu, K.; Frantz, S.; Jaffer, F.A.; Tung, C.H.; Hiller, K.H.; Voll, S.; Nordbeck, P.; Sosnovik, D.; Gattenlohner, S., et al. Factor xiii deficiency causes cardiac rupture, impairs wound healing, and aggravates cardiac remodeling in mice with myocardial infarction. Circulation 2006, 113, 1196-1202. 263. Inbal, A.; Lubetsky, A.; Krapp, T.; Castel, D.; Shaish, A.; Dickneitte, G.; Modis, L.; Muszbek, L.; Inbal, A. Impaired wound healing in factor xiii deficient mice. Thrombosis and haemostasis 2005, 94, 432-437. 264. Dardik, R.; Leor, J.; Skutelsky, E.; Castel, D.; Holbova, R.; Schiby, G.; Shaish, A.; Dickneite, G.; Loscalzo, J.; Inbal, A. Evaluation of the pro-angiogenic effect of factor xiii in heterotopic mouse heart allografts and fxiii-deficient mice. Thrombosis and haemostasis 2006, 95, 546-550. 265. Loof, T.G.; Morgelin, M.; Johansson, L.; Oehmcke, S.; Olin, A.I.; Dickneite, G.; Norrby-Teglund, A.; Theopold, U.; Herwald, H. Coagulation, an ancestral serine protease cascade, exerts a novel function in early immune defense. Blood 2011, 118, 2589-2598. 266. Klock, C.; Khosla, C. Regulation of the activities of the mammalian transglutaminase family of enzymes. Protein science : a publication of the Protein Society 2012, 21, 1781-1791. 267. Steinert, P.M.; Kim, S.Y.; Chung, S.I.; Marekov, L.N. The transglutaminase 1 enzyme is variably acylated by myristate and palmitate during differentiation in epidermal keratinocytes. The Journal of biological chemistry 1996, 271, 26242-26250. 268. Steinert, P.M.; Chung, S.I.; Kim, S.Y. Inactive zymogen and highly active proteolytically processed membrane-bound forms of the transglutaminase 1 enzyme in human epidermal keratinocytes. Biochemical and biophysical research communications 1996, 221, 101-106. 269. Kim, S.Y.; Chung, S.I.; Steinert, P.M. Highly active soluble processed forms of the transglutaminase 1 enzyme in epidermal keratinocytes. The Journal of biological chemistry 1995, 270, 18026-18035. 270. Iizuka, R.; Chiba, K.; Imajoh-Ohmi, S. A novel approach for the detection of proteolytically activated transglutaminase 1 in epidermis using cleavage site-directed antibodies. The Journal of investigative dermatology 2003, 121, 457-464. 271. Egberts, F.; Heinrich, M.; Jensen, J.M.; Winoto-Morbach, S.; Pfeiffer, S.; Wickel, M.; Schunck, M.; Steude, J.; Saftig, P.; Proksch, E., et al. Cathepsin d is involved in the regulation of transglutaminase 1 and epidermal differentiation. Journal of cell science 2004, 117, 2295-2307. 272. Boeshans, K.M.; Mueser, T.C.; Ahvazi, B. A three-dimensional model of the human transglutaminase 1: Insights into the understanding of lamellar ichthyosis. Journal of molecular modeling 2007, 13, 233-246. 273. Matsuki, M.; Yamashita, F.; Ishida-Yamamoto, A.; Yamada, K.; Kinoshita, C.; Fushiki, S.; Ueda, E.; Morishima, Y.; Tabata, K.; Yasuno, H., et al. Defective stratum corneum and early neonatal death in mice lacking the gene for transglutaminase 1 (keratinocyte transglutaminase). Proceedings of the National Academy of Sciences of the United States of America 1998, 95, 1044-1049. 274. John, S.; Thiebach, L.; Frie, C.; Mokkapati, S.; Bechtel, M.; Nischt, R.; Rosser-Davies, S.; Paulsson, M.; Smyth, N. Epidermal transglutaminase (tgase 3) is required for proper hair development, but not the formation of the epidermal barrier. PloS one 2012, 7, e34252. 275. Dean, M.D. Genetic disruption of the copulatory plug in mice leads to severely reduced fertility. PLoS genetics 2013, 9, e1003185. 276. Ablin, R.J.; Owen, S.; Jiang, W.G. Prostate transglutaminase (tgase-4) induces epithelial-to- mesenchymal transition in prostate cancer cells. Anticancer research 2017, 37, 481-487. 277. Cassidy, A.J.; van Steensel, M.A.; Steijlen, P.M.; van Geel, M.; van der Velden, J.; Morley, S.M.; Terrinoni, A.; Melino, G.; Candi, E.; McLean, W.H. A homozygous missense mutation in tgm5 abolishes epidermal transglutaminase 5 activity and causes acral peeling skin syndrome. American journal of human genetics 2005, 77, 909-917.

144

278. Candi, E.; Oddi, S.; Paradisi, A.; Terrinoni, A.; Ranalli, M.; Teofoli, P.; Citro, G.; Scarpato, S.; Puddu, P.; Melino, G. Expression of transglutaminase 5 in normal and pathologic human epidermis. The Journal of investigative dermatology 2002, 119, 670-677. 279. Tripathy, D.; Vignoli, B.; Ramesh, N.; Polanco, M.J.; Coutelier, M.; Stephen, C.D.; Canossa, M.; Monin, M.L.; Aeschlimann, P.; Turberville, S., et al. Mutations in tgm6 induce the unfolded protein response in sca35. Human molecular genetics 2017, 26, 3749-3762. 280. Nurminskaya, M.; Kaartinen, M.T. Transglutaminases in mineralized tissues. Frontiers in bioscience : a journal and virtual library 2006, 11, 1591-1606. 281. Al-Jallad, H.F.; Nakano, Y.; Chen, J.L.; McMillan, E.; Lefebvre, C.; Kaartinen, M.T. Transglutaminase activity regulates osteoblast differentiation and matrix mineralization in mc3t3-e1 osteoblast cultures. Matrix biology : journal of the International Society for Matrix Biology 2006, 25, 135-148. 282. Forsprecher, J.; Wang, Z.; Nelea, V.; Kaartinen, M.T. Enhanced osteoblast adhesion on transglutaminase 2-crosslinked fibronectin. Amino acids 2009, 36, 747-753. 283. Forsprecher, J.; Wang, Z.; Goldberg, H.A.; Kaartinen, M.T. Transglutaminase-mediated oligomerization promotes osteoblast adhesive properties of osteopontin and bone sialoprotein. Cell adhesion & migration 2011, 5, 65-72. 284. Piercy-Kotb, S.A.; Mousa, A.; Al-Jallad, H.F.; Myneni, V.D.; Chicatun, F.; Nazhat, S.N.; Kaartinen, M.T. Factor xiiia transglutaminase expression and secretion by osteoblasts is regulated by extracellular matrix collagen and the map kinase signaling pathway. Journal of cellular physiology 2012, 227, 2936-2946. 285. Cui, C.; Wang, S.; Myneni, V.D.; Hitomi, K.; Kaartinen, M.T. Transglutaminase activity arising from factor xiiia is required for stabilization and conversion of plasma fibronectin into matrix in osteoblast cultures. Bone 2014, 59, 127-138. 286. Al-Jallad, H.F.; Myneni, V.D.; Piercy-Kotb, S.A.; Chabot, N.; Mulani, A.; Keillor, J.W.; Kaartinen, M.T. Plasma membrane factor xiiia transglutaminase activity regulates osteoblast matrix secretion and deposition by affecting microtubule dynamics. PloS one 2011, 6, e15893. 287. Lauer, P.; Metzner, H.J.; Zettlmeissl, G.; Li, M.; Smith, A.G.; Lathe, R.; Dickneite, G. Targeted inactivation of the mouse locus encoding coagulation factor xiii-a: Hemostatic abnormalities in mutant mice and characterization of the coagulation deficit. Thrombosis and haemostasis 2002, 88, 967-974. 288. Mousa, A.; Cui, C.; Song, A.; Myneni, V.D.; Sun, H.; Li, J.J.; Murshed, M.; Melino, G.; Kaartinen, M.T. Transglutaminases factor xiii-a and tg2 regulate resorption, adipogenesis and plasma fibronectin homeostasis in bone and bone marrow. Cell death and differentiation 2017, 24, 844-854. 289. Raghu, H.; Cruz, C.; Rewerts, C.L.; Frederick, M.D.; Thornton, S.; Mullins, E.S.; Schoenecker, J.G.; Degen, J.L.; Flick, M.J. Transglutaminase factor xiii promotes arthritis through mechanisms linked to inflammation and bone erosion. Blood 2015, 125, 427-437. 290. Kim, W.S.; Kim, H.; Jeong, E.M.; Kim, H.J.; Lee, Z.H.; Kim, I.G.; Kim, H.H. Transglutaminase 2 regulates osteoclast differentiation via a blimp1-dependent pathway. Scientific reports 2017, 7, 10626. 291. Sun, H.; Kaartinen, M.T. Transglutaminase activity regulates differentiation, migration and fusion of osteoclasts via affecting actin dynamics. Journal of cellular physiology 2018, 233, 7497-7513. 292. Haldar, M.; Murphy, K.M. Origin, development, and homeostasis of tissue-resident macrophages. Immunological reviews 2014, 262, 25-35. 293. Gordon, S.; Pluddemann, A. Macrophage clearance of apoptotic cells: A critical assessment. Frontiers in immunology 2018, 9, 127. 294. Jou, I.M.; Lin, C.F.; Tsai, K.J.; Wei, S.J. Macrophage-mediated inflammatory disorders. Mediators of inflammation 2013, 2013, 316482.

145

295. Bagoly, Z.; Katona, E.; Muszbek, L. Factor xiii and inflammatory cells. Thrombosis research 2012, 129 Suppl 2, S77-81. 296. Chrobok, N.L.; Sestito, C.; Wilhelmus, M.M.; Drukarch, B.; van Dam, A.M. Is monocyte- and macrophage-derived tissue transglutaminase involved in inflammatory processes? Amino acids 2017, 49, 441-452. 297. Gundemir, S.; Colak, G.; Tucholski, J.; Johnson, G.V. Transglutaminase 2: A molecular swiss army knife. Biochimica et biophysica acta 2012, 1823, 406-419. 298. Myneni, V.D.; Hitomi, K.; Kaartinen, M.T. Factor xiii-a transglutaminase acts as a switch between preadipocyte proliferation and differentiation. Blood 2014, 124, 1344-1353. 299. Huber, M.; Rettler, I.; Bernasconi, K.; Frenk, E.; Lavrijsen, S.P.; Ponec, M.; Bon, A.; Lautenschlager, S.; Schorderet, D.F.; Hohl, D. Mutations of keratinocyte transglutaminase in lamellar ichthyosis. Science (New York, N.Y.) 1995, 267, 525-528. 300. Parmentier, L.; Blanchet-Bardon, C.; Nguyen, S.; Prud'homme, J.F.; Dubertret, L.; Weissenbach, J. Autosomal recessive lamellar ichthyosis: Identification of a new mutation in transglutaminase 1 and evidence for genetic heterogeneity. Human molecular genetics 1995, 4, 1391-1395. 301. Russell, L.J.; DiGiovanna, J.J.; Rogers, G.R.; Steinert, P.M.; Hashem, N.; Compton, J.G.; Bale, S.J. Mutations in the gene for transglutaminase 1 in autosomal recessive lamellar ichthyosis. Nature genetics 1995, 9, 279-283. 302. Bakker, E.N.; Pistea, A.; Spaan, J.A.; Rolf, T.; de Vries, C.J.; van Rooijen, N.; Candi, E.; VanBavel, E. Flow-dependent remodeling of small arteries in mice deficient for tissue-type transglutaminase: Possible compensation by macrophage-derived factor xiii. Circulation research 2006, 99, 86-92. 303. Karpati, S.; Sardy, M.; Nemeth, K.; Mayer, B.; Smyth, N.; Paulsson, M.; Traupe, H. Transglutaminases in autoimmune and inherited skin diseases: The phenomena of epitope spreading and functional compensation. Experimental dermatology 2017. 304. Odii, B.O.; Coussons, P. Biological functionalities of transglutaminase 2 and the possibility of its compensation by other members of the transglutaminase family. TheScientificWorldJournal 2014, 2014, 714561. 305. Deasey, S.; Shanmugasundaram, S.; Nurminskaya, M. Tissue-specific responses to loss of transglutaminase 2. Amino acids 2013, 44, 179-187. 306. van Furth, R.; Diesselhoff-den Dulk, M.M. Dual origin of mouse spleen macrophages. The Journal of experimental medicine 1984, 160, 1273-1283. 307. van Furth, R.; Cohn, Z.A. The origin and kinetics of mononuclear phagocytes. The Journal of experimental medicine 1968, 128, 415-435. 308. Ginhoux, F.; Guilliams, M. Tissue-resident macrophage ontogeny and homeostasis. Immunity 2016, 44, 439-449. 309. Ginhoux, F.; Greter, M.; Leboeuf, M.; Nandi, S.; See, P.; Gokhan, S.; Mehler, M.F.; Conway, S.J.; Ng, L.G.; Stanley, E.R., et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages. Science (New York, N.Y.) 2010, 330, 841-845. 310. Hoeffel, G.; Wang, Y.; Greter, M.; See, P.; Teo, P.; Malleret, B.; Leboeuf, M.; Low, D.; Oller, G.; Almeida, F., et al. Adult langerhans cells derive predominantly from embryonic fetal liver monocytes with a minor contribution of yolk sac-derived macrophages. The Journal of experimental medicine 2012, 209, 1167- 1181. 311. Schulz, C.; Gomez Perdiguero, E.; Chorro, L.; Szabo-Rogers, H.; Cagnard, N.; Kierdorf, K.; Prinz, M.; Wu, B.; Jacobsen, S.E.; Pollard, J.W., et al. A lineage of myeloid cells independent of myb and

146

hematopoietic stem cells. Science (New York, N.Y.) 2012, 336, 86-90. 312. Epelman, S.; Lavine, K.J.; Randolph, G.J. Origin and functions of tissue macrophages. Immunity 2014, 41, 21-35. 313. Molawi, K.; Wolf, Y.; Kandalla, P.K.; Favret, J.; Hagemeyer, N.; Frenzel, K.; Pinto, A.R.; Klapproth, K.; Henri, S.; Malissen, B., et al. Progressive replacement of embryo-derived cardiac macrophages with age. The Journal of experimental medicine 2014, 211, 2151-2158. 314. Gomez Perdiguero, E.; Klapproth, K.; Schulz, C.; Busch, K.; Azzoni, E.; Crozet, L.; Garner, H.; Trouillet, C.; de Bruijn, M.F.; Geissmann, F., et al. Tissue-resident macrophages originate from yolk-sac-derived erythro-myeloid progenitors. Nature 2015, 518, 547-551. 315. Ajami, B.; Bennett, J.L.; Krieger, C.; Tetzlaff, W.; Rossi, F.M. Local self-renewal can sustain cns microglia maintenance and function throughout adult life. Nature neuroscience 2007, 10, 1538-1543. 316. Mildner, A.; Schmidt, H.; Nitsche, M.; Merkler, D.; Hanisch, U.K.; Mack, M.; Heikenwalder, M.; Bruck, W.; Priller, J.; Prinz, M. Microglia in the adult brain arise from ly-6chiccr2+ monocytes only under defined host conditions. Nature neuroscience 2007, 10, 1544-1553. 317. Hashimoto, D.; Chow, A.; Noizat, C.; Teo, P.; Beasley, M.B.; Leboeuf, M.; Becker, C.D.; See, P.; Price, J.; Lucas, D., et al. Tissue-resident macrophages self-maintain locally throughout adult life with minimal contribution from circulating monocytes. Immunity 2013, 38, 792-804. 318. Sieweke, M.H.; Allen, J.E. Beyond stem cells: Self-renewal of differentiated macrophages. Science (New York, N.Y.) 2013, 342, 1242974. 319. Yona, S.; Kim, K.W.; Wolf, Y.; Mildner, A.; Varol, D.; Breker, M.; Strauss-Ayali, D.; Viukov, S.; Guilliams, M.; Misharin, A., et al. Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeostasis. Immunity 2013, 38, 79-91. 320. Mills, C.D. M1 and m2 macrophages: Oracles of health and disease. Critical reviews in immunology 2012, 32, 463-488. 321. Murtaugh, M.P.; Arend, W.P.; Davies, P.J. Induction of tissue transglutaminase in human peripheral blood monocytes. The Journal of experimental medicine 1984, 159, 114-125. 322. Muszbek, L.; Adany, R.; Szegedi, G.; Polgar, J.; Kavai, M. Factor xiii of blood coagulation in human monocytes. Thrombosis research 1985, 37, 401-410. 323. Akimov, S.S.; Belkin, A.M. Cell surface tissue transglutaminase is involved in adhesion and migration of monocytic cells on fibronectin. Blood 2001, 98, 1567-1576. 324. Thomas-Ecker, S.; Lindecke, A.; Hatzmann, W.; Kaltschmidt, C.; Zanker, K.S.; Dittmar, T. Alteration in the gene expression pattern of primary monocytes after adhesion to endothelial cells. Proceedings of the National Academy of Sciences of the United States of America 2007, 104, 5539-5544. 325. AbdAlla, S.; Lother, H.; Langer, A.; el Faramawy, Y.; Quitterer, U. Factor xiiia transglutaminase crosslinks at1 receptor dimers of monocytes at the onset of atherosclerosis. Cell 2004, 119, 343-354. 326. Schroff, G.; Neumann, C.; Sorg, C. Transglutaminase as a marker for subsets of murine macrophages. European journal of immunology 1981, 11, 637-642. 327. Adany, R.; Belkin, A.; Vasilevskaya, T.; Muszbek, L. Identification of blood coagulation factor xiii in human peritoneal macrophages. European journal of cell biology 1985, 38, 171-173. 328. Henriksson, P.; Becker, S.; Lynch, G.; McDonagh, J. Identification of intracellular factor xiii in human monocytes and macrophages. The Journal of clinical investigation 1985, 76, 528-534. 329. Murtaugh, M.P.; Mehta, K.; Johnson, J.; Myers, M.; Juliano, R.L.; Davies, P.J. Induction of tissue transglutaminase in mouse peritoneal macrophages. The Journal of biological chemistry 1983, 258, 11074-11081.

147

330. Torocsik, D.; Bardos, H.; Nagy, L.; Adany, R. Identification of factor xiii-a as a marker of alternative macrophage activation. Cellular and molecular life sciences : CMLS 2005, 62, 2132-2139. 331. Martinez, F.O.; Helming, L.; Milde, R.; Varin, A.; Melgert, B.N.; Draijer, C.; Thomas, B.; Fabbri, M.; Crawshaw, A.; Ho, L.P., et al. Genetic programs expressed in resting and il-4 alternatively activated mouse and human macrophages: Similarities and differences. Blood 2013, 121, e57-69. 332. Mehta, K.; Lopez-Berestein, G. Expression of tissue transglutaminase in cultured monocytic leukemia (thp-1) cells during differentiation. Cancer research 1986, 46, 1388-1394. 333. Metha, K.; Turpin, J.; Lopez-Berestein, G. Induction of tissue transglutaminase in human peripheral blood monocytes by intracellular delivery of retinoids. Journal of leukocyte biology 1987, 41, 341-348. 334. Mehta, K.; Lopez-Berestein, G.; Moore, W.T.; Davies, P.J. Interferon-gamma requires serum retinoids to promote the expression of tissue transglutaminase in cultured human blood monocytes. Journal of immunology (Baltimore, Md. : 1950) 1985, 134, 2053-2056. 335. Chiocca, E.A.; Davies, P.J.; Stein, J.P. Regulation of tissue transglutaminase gene expression as a molecular model for retinoid effects on proliferation and differentiation. Journal of cellular biochemistry 1989, 39, 293-304. 336. Moore, W.T., Jr.; Murtaugh, M.P.; Davies, P.J. Retinoic acid-induced expression of tissue transglutaminase in mouse peritoneal macrophages. The Journal of biological chemistry 1984, 259, 12794-12802. 337. Chiocca, E.A.; Davies, P.J.; Stein, J.P. The molecular basis of retinoic acid action. Transcriptional regulation of tissue transglutaminase gene expression in macrophages. The Journal of biological chemistry 1988, 263, 11584-11589. 338. Murtaugh, M.P.; Dennison, O.; Stein, J.P.; Davies, P.J. Retinoic acid-induced gene expression in normal and leukemic myeloid cells. The Journal of experimental medicine 1986, 163, 1325-1330. 339. Murtaugh, M.P.; Moore, W.T., Jr.; Davies, P.J. Cyclic amp potentiates the retinoic acid-induced expression of tissue transglutaminase in peritoneal macrophages. The Journal of biological chemistry 1986, 261, 614-621. 340. Ishii, I.; Ui, M. Retinoic acid-induced gene expression of tissue transglutaminase via protein kinase c- dependent pathway in mouse peritoneal macrophages. Journal of biochemistry 1994, 115, 1197-1202. 341. Ghanta, K.S.; Pakala, S.B.; Reddy, S.D.; Li, D.Q.; Nair, S.S.; Kumar, R. Mta1 coregulation of transglutaminase 2 expression and function during inflammatory response. The Journal of biological chemistry 2011, 286, 7132-7138. 342. Falasca, L.; Farrace, M.G.; Rinaldi, A.; Tuosto, L.; Melino, G.; Piacentini, M. Transglutaminase type ii is involved in the pathogenesis of endotoxic shock. Journal of immunology (Baltimore, Md. : 1950) 2008, 180, 2616-2624. 343. Curro, M.; Ferlazzo, N.; Risitano, R.; Condello, S.; Vecchio, M.; Caccamo, D.; Ientile, R. Transglutaminase 2 and phospholipase a(2) interactions in the inflammatory response in human thp-1 monocytes. Amino acids 2014, 46, 759-766. 344. Sarang, Z.; Koroskenyi, K.; Pallai, A.; Duro, E.; Melino, G.; Griffin, M.; Fesus, L.; Szondy, Z. Transglutaminase 2 null macrophages respond to lipopolysaccharide stimulation by elevated proinflammatory cytokine production due to an enhanced alphavbeta3 integrin-induced src tyrosine kinase signaling. Immunology letters 2011, 138, 71-78. 345. Yoo, H.; Ahn, E.R.; Kim, S.J.; Lee, S.H.; Oh, S.H.; Kim, S.Y. Divergent results induced by different types of septic shock in transglutaminase 2 knockout mice. Amino acids 2013, 44, 189-197. 346. Gratchev, A.; Kzhyshkowska, J.; Utikal, J.; Goerdt, S. Interleukin-4 and dexamethasone counterregulate

148

extracellular matrix remodelling and phagocytosis in type-2 macrophages. Scandinavian journal of immunology 2005, 61, 10-17. 347. Yamaguchi, M.; Zacharia, J.; Laidlaw, T.M.; Balestrieri, B. Pla2g5 regulates transglutaminase activity of human il-4-activated m2 macrophages through pge2 generation. Journal of leukocyte biology 2016, 100, 131-141. 348. Ribas, V.; Drew, B.G.; Le, J.A.; Soleymani, T.; Daraei, P.; Sitz, D.; Mohammad, L.; Henstridge, D.C.; Febbraio, M.A.; Hewitt, S.C., et al. Myeloid-specific estrogen receptor alpha deficiency impairs metabolic homeostasis and accelerates atherosclerotic lesion development. Proceedings of the National Academy of Sciences of the United States of America 2011, 108, 16457-16462. 349. Frisdal, E.; Lesnik, P.; Olivier, M.; Robillard, P.; Chapman, M.J.; Huby, T.; Guerin, M.; Le Goff, W. Interleukin-6 protects human macrophages from cellular cholesterol accumulation and attenuates the proinflammatory response. The Journal of biological chemistry 2011, 286, 30926-30936. 350. Yen, J.H.; Lin, L.C.; Chen, M.C.; Sarang, Z.; Leong, P.Y.; Chang, I.C.; Hsu, J.D.; Chen, J.H.; Hsieh, Y.F.; Pallai, A., et al. The metastatic tumor antigen 1-transglutaminase-2 pathway is involved in self-limitation of monosodium urate crystal-induced inflammation by upregulating tgf-beta1. Arthritis research & therapy 2015, 17, 65. 351. Shinde, A.V.; Dobaczewski, M.; de Haan, J.J.; Saxena, A.; Lee, K.K.; Xia, Y.; Chen, W.; Su, Y.; Hanif, W.; Kaur Madahar, I., et al. Tissue transglutaminase induction in the pressure-overloaded myocardium regulates matrix remodelling. Cardiovascular research 2017, 113, 892-905. 352. Seiving, B.; Ohlsson, K.; Linder, C.; Stenberg, P. Transglutaminase differentiation during maturation of human blood monocytes to macrophages. European journal of haematology 1991, 46, 263-271. 353. Eligini, S.; Fiorelli, S.; Tremoli, E.; Colli, S. Inhibition of transglutaminase 2 reduces efferocytosis in human macrophages: Role of cd14 and sr-ai receptors. Nutrition, metabolism, and cardiovascular diseases : NMCD 2016, 26, 922-930. 354. Sarvary, A.; Szucs, S.; Balogh, I.; Becsky, A.; Bardos, H.; Kavai, M.; Seligsohn, U.; Egbring, R.; Lopaciuk, S.; Muszbek, L., et al. Possible role of factor xiii subunit a in fcgamma and complement receptor-mediated phagocytosis. Cellular immunology 2004, 228, 81-90. 355. Conkling, P.R.; Achyuthan, K.E.; Greenberg, C.S.; Newcomb, T.F.; Weinberg, J.B. Human mononuclear phagocyte transglutaminase activity cross-links fibrin. Thrombosis research 1989, 55, 57-68. 356. Adany, R.; Bardos, H.; Antal, M.; Modis, L.; Sarvary, A.; Szucs, S.; Balogh, I. Factor xiii of blood coagulation as a nuclear crosslinking enzyme. Thrombosis and haemostasis 2001, 85, 845-851. 357. Scheller, J.; Chalaris, A.; Schmidt-Arras, D.; Rose-John, S. The pro- and anti-inflammatory properties of the cytokine interleukin-6. Biochimica et biophysica acta 2011, 1813, 878-888. 358. Sanjabi, S.; Zenewicz, L.A.; Kamanaka, M.; Flavell, R.A. Anti-inflammatory and pro-inflammatory roles of tgf-beta, il-10, and il-22 in immunity and autoimmunity. Current opinion in pharmacology 2009, 9, 447-453. 359. Bohdanowicz, M.; Grinstein, S. Role of phospholipids in endocytosis, phagocytosis, and macropinocytosis. Physiological reviews 2013, 93, 69-106. 360. Vernon, P.J.; Tang, D. Eat-me: Autophagy, phagocytosis, and reactive oxygen species signaling. Antioxidants & redox signaling 2013, 18, 677-691. 361. Korns, D.; Frasch, S.C.; Fernandez-Boyanapalli, R.; Henson, P.M.; Bratton, D.L. Modulation of macrophage efferocytosis in inflammation. Frontiers in immunology 2011, 2, 57. 362. Fesus, L.; Sandor, M.; Horvath, L.I.; Bagyinka, C.; Erdei, A.; Gergely, J. Immune-complex-induced transglutaminase activation: Its role in the fc-receptor-mediated transmembrane effect on peritoneal

149

macrophages. Molecular immunology 1981, 18, 633-638. 363. Leu, R.W.; Herriott, M.J.; Moore, P.E.; Orr, G.R.; Birckbichler, P.J. Enhanced transglutaminase activity associated with macrophage activation. Possible role in fc-mediated phagocytosis. Experimental cell research 1982, 141, 191-199. 364. Abe, S.; Yamashita, K.; Kohno, H.; Ohkubo, Y. Involvement of transglutaminase in the receptor- mediated endocytosis of mouse peritoneal macrophages. Biological & pharmaceutical bulletin 2000, 23, 1511-1513. 365. Rose, D.M.; Sydlaske, A.D.; Agha-Babakhani, A.; Johnson, K.; Terkeltaub, R. Transglutaminase 2 limits murine peritoneal acute gout-like inflammation by regulating macrophage clearance of apoptotic neutrophils. Arthritis and rheumatism 2006, 54, 3363-3371. 366. Boisvert, W.A.; Rose, D.M.; Boullier, A.; Quehenberger, O.; Sydlaske, A.; Johnson, K.A.; Curtiss, L.K.; Terkeltaub, R. Leukocyte transglutaminase 2 expression limits atherosclerotic lesion size. Arteriosclerosis, thrombosis, and vascular biology 2006, 26, 563-569. 367. Fadok, V.A.; Bratton, D.L.; Konowal, A.; Freed, P.W.; Westcott, J.Y.; Henson, P.M. Macrophages that have ingested apoptotic cells in vitro inhibit proinflammatory cytokine production through autocrine/paracrine mechanisms involving tgf-beta, pge2, and paf. The Journal of clinical investigation 1998, 101, 890-898. 368. Kojima, S.; Nara, K.; Rifkin, D.B. Requirement for transglutaminase in the activation of latent transforming growth factor-beta in bovine endothelial cells. The Journal of cell biology 1993, 121, 439- 448. 369. Toth, B.; Garabuczi, E.; Sarang, Z.; Vereb, G.; Vamosi, G.; Aeschlimann, D.; Blasko, B.; Becsi, B.; Erdodi, F.; Lacy-Hulbert, A., et al. Transglutaminase 2 is needed for the formation of an efficient phagocyte portal in macrophages engulfing apoptotic cells. Journal of immunology (Baltimore, Md. : 1950) 2009, 182, 2084-2092. 370. Toth, B.; Sarang, Z.; Vereb, G.; Zhang, A.; Tanaka, S.; Melino, G.; Fesus, L.; Szondy, Z. Over-expression of integrin beta3 can partially overcome the defect of integrin beta3 signaling in transglutaminase 2 null macrophages. Immunology letters 2009, 126, 22-28. 371. Nadella, V.; Wang, Z.; Johnson, T.S.; Griffin, M.; Devitt, A. Transglutaminase 2 interacts with syndecan- 4 and cd44 at the surface of human macrophages to promote removal of apoptotic cells. Biochimica et biophysica acta 2015, 1853, 201-212. 372. Sarang, Z.; Joos, G.; Garabuczi, E.; Ruhl, R.; Gregory, C.D.; Szondy, Z. Macrophages engulfing apoptotic cells produce nonclassical retinoids to enhance their phagocytic capacity. Journal of immunology (Baltimore, Md. : 1950) 2014, 192, 5730-5738. 373. Kavai, M.; Adany, R.; Pasti, G.; Suranyi, P.; Szucs, G.; Muszbek, L.; Bojan, F.; Szegedi, G. Marker profile, enzyme activity, and function of a human myelomonocytic leukemia cell line. Cellular immunology 1992, 139, 531-540. 374. Tousoulis, D.; Kampoli, A.M.; Papageorgiou, N.; Androulakis, E.; Antoniades, C.; Toutouzas, K.; Stefanadis, C. Pathophysiology of atherosclerosis: The role of inflammation. Current pharmaceutical design 2011, 17, 4089-4110. 375. Ley, K.; Miller, Y.I.; Hedrick, C.C. Monocyte and macrophage dynamics during atherogenesis. Arteriosclerosis, thrombosis, and vascular biology 2011, 31, 1506-1516. 376. Haroon, Z.A.; Wannenburg, T.; Gupta, M.; Greenberg, C.S.; Wallin, R.; Sane, D.C. Localization of tissue transglutaminase in human carotid and coronary artery atherosclerosis: Implications for plaque stability and progression. Laboratory investigation; a journal of technical methods and pathology 2001, 81, 83-

150

93. 377. Sumi, Y.; Inoue, N.; Azumi, H.; Seno, T.; Okuda, M.; Hirata, K.; Kawashima, S.; Hayashi, Y.; Itoh, H.; Yokoyama, M. Expression of tissue transglutaminase and elafin in human coronary artery: Implication for plaque instability. Atherosclerosis 2002, 160, 31-39. 378. Romanic, A.M.; Arleth, A.J.; Willette, R.N.; Ohlstein, E.H. Factor xiiia cross-links lipoprotein(a) with fibrinogen and is present in human atherosclerotic lesions. Circulation research 1998, 83, 264-269. 379. Matlung, H.L.; Groen, H.C.; de Vos, J.; van Walsum, T.; van der Lugt, A.; Niessen, W.J.; Wentzel, J.J.; Vanbavel, E.; Bakker, E.N. Calcification locates to transglutaminases in advanced human atherosclerotic lesions. The American journal of pathology 2009, 175, 1374-1379. 380. Van Herck, J.L.; Schrijvers, D.M.; De Meyer, G.R.; Martinet, W.; Van Hove, C.E.; Bult, H.; Vrints, C.J.; Herman, A.G. Transglutaminase 2 deficiency decreases plaque fibrosis and increases plaque inflammation in apolipoprotein-e-deficient mice. Journal of vascular research 2010, 47, 231-240. 381. Matlung, H.L.; Neele, A.E.; Groen, H.C.; van Gaalen, K.; Tuna, B.G.; van Weert, A.; de Vos, J.; Wentzel, J.J.; Hoogenboezem, M.; van Buul, J.D., et al. Transglutaminase activity regulates atherosclerotic plaque composition at locations exposed to oscillatory shear stress. Atherosclerosis 2012, 224, 355-362. 382. Matlung, H.L.; VanBavel, E.; van den Akker, J.; de Vries, C.J.; Bakker, E.N. Role of transglutaminases in cuff-induced atherosclerotic lesion formation in femoral arteries of apoe3 leiden mice. Atherosclerosis 2010, 213, 77-84. 383. Tan, S.Y.; Krasnow, M.A. Developmental origin of lung macrophage diversity. Development (Cambridge, England) 2016, 143, 1318-1327. 384. Roth, W.J.; Chung, S.I.; Janoff, A. Inactivation of alveolar macrophage transglutaminase by oxidants in cigarette smoke. Journal of leukocyte biology 1986, 39, 629-644. 385. Roth, W.J.; Fleit, H.B.; Chung, S.I.; Janoff, A. Characterization of two distinct transglutaminases of murine bone marrow-derived macrophages: Effects of exposure of viable cells to cigarette smoke on enzyme activity. Journal of leukocyte biology 1987, 42, 9-20. 386. Roth, W.J.; Chung, S.I.; Raju, L.; Janoff, A. Macrophage transglutaminases: Characterization of molecular species and measurement of enzymatic modification by cigarette smoke components. Advances in experimental medicine and biology 1988, 231, 161-173. 387. Dietz, K.; de Los Reyes Jimenez, M.; Gollwitzer, E.S.; Chaker, A.M.; Zissler, U.M.; Radmark, O.P.; Baarsma, H.A.; Konigshoff, M.; Schmidt-Weber, C.B.; Marsland, B.J., et al. Age dictates a steroid- resistant cascade of wnt5a, transglutaminase 2, and leukotrienes in inflamed airways. The Journal of allergy and clinical immunology 2017, 139, 1343-1354.e1346. 388. Kradin, R.L.; Lynch, G.W.; Kurnick, J.T.; Erikson, M.; Colvin, R.B.; McDonagh, J. Factor xiii a is synthesized and expressed on the surface of u937 cells and alveolar macrophages. Blood 1987, 69, 778- 785. 389. Katona, E.; Nagy, B.; Kappelmayer, J.; Baktai, G.; Kovacs, L.; Marialigeti, T.; Dezso, B.; Muszbek, L. Factor xiii in bronchoalveolar lavage fluid from children with chronic bronchoalveolar inflammation. Journal of thrombosis and haemostasis : JTH 2005, 3, 1407-1413. 390. Nie, H.; Wang, A.; He, Q.; Yang, Q.; Liu, L.; Zhang, G.; Huang, Y.; Ding, X.; Yu, H.; Hu, S. Phenotypic switch in lung interstitial macrophage polarization in an ovalbumin-induced mouse model of asthma. Experimental and therapeutic medicine 2017, 14, 1284-1292. 391. Sousa, C.; Biber, K.; Michelucci, A. Cellular and molecular characterization of microglia: A unique immune cell population. Frontiers in immunology 2017, 8, 198. 392. Lannes, N.; Eppler, E.; Etemad, S.; Yotovski, P.; Filgueira, L. Microglia at center stage: A comprehensive

151

review about the versatile and unique residential macrophages of the central nervous system. Oncotarget 2017, 8, 114393-114413. 393. Boche, D.; Perry, V.H.; Nicoll, J.A. Review: Activation patterns of microglia and their identification in the human brain. Neuropathology and applied neurobiology 2013, 39, 3-18. 394. Neniskyte, U.; Neher, J.J.; Brown, G.C. Neuronal death induced by nanomolar amyloid beta is mediated by primary phagocytosis of neurons by microglia. The Journal of biological chemistry 2011, 286, 39904- 39913. 395. Contestabile, A.; Monti, B.; Polazzi, E. Neuronal-glial interactions define the role of nitric oxide in neural functional processes. Current neuropharmacology 2012, 10, 303-310. 396. Park, K.C.; Chung, K.C.; Kim, Y.S.; Lee, J.; Joh, T.H.; Kim, S.Y. Transglutaminase 2 induces nitric oxide synthesis in bv-2 microglia. Biochemical and biophysical research communications 2004, 323, 1055- 1062. 397. Kawabe, K.; Takano, K.; Moriyama, M.; Nakamura, Y. Lipopolysaccharide-stimulated transglutaminase 2 expression enhances endocytosis activity in the mouse microglial cell line bv-2. Neuroimmunomodulation 2015, 22, 243-249. 398. Lee, J.; Kim, Y.S.; Choi, D.H.; Bang, M.S.; Han, T.R.; Joh, T.H.; Kim, S.Y. Transglutaminase 2 induces nuclear factor-kappab activation via a novel pathway in bv-2 microglia. The Journal of biological chemistry 2004, 279, 53725-53735. 399. Kawabe, K.; Takano, K.; Moriyama, M.; Nakamura, Y. Amphotericin b increases transglutaminase 2 expression associated with upregulation of endocytotic activity in mouse microglial cell line bv-2. Neurochemical research 2017, 42, 1488-1495. 400. Kawabe, K.; Takano, K.; Moriyama, M.; Nakamura, Y. Microglia endocytose amyloid beta through the binding of transglutaminase 2 and milk fat globule egf factor 8 protein. Neurochemical research 2018, 43, 32-40. 401. Curro, M.; Ferlazzo, N.; Condello, S.; Caccamo, D.; Ientile, R. Transglutaminase 2 silencing reduced the beta-amyloid-effects on the activation of human thp-1 cells. Amino acids 2010, 39, 1427-1433. 402. Curro, M.; Gangemi, C.; Giunta, M.L.; Ferlazzo, N.; Navarra, M.; Ientile, R.; Caccamo, D. Transglutaminase 2 is involved in amyloid-beta1-42-induced pro-inflammatory activation via ap1/jnk signalling pathways in thp-1 monocytes. Amino acids 2017, 49, 659-669. 403. Yamada, T.; Yoshiyama, Y.; Kawaguchi, N.; Ichinose, A.; Iwaki, T.; Hirose, S.; Jefferies, W.A. Possible roles of transglutaminases in alzheimer's disease. Dementia and geriatric cognitive disorders 1998, 9, 103-110. 404. Noseworthy, J.H.; Lucchinetti, C.; Rodriguez, M.; Weinshenker, B.G. Multiple sclerosis. The New England journal of medicine 2000, 343, 938-952. 405. Koch-Henriksen, N.; Sorensen, P.S. The changing demographic pattern of multiple sclerosis epidemiology. The Lancet. Neurology 2010, 9, 520-532. 406. Stadelmann, C.; Wegner, C.; Bruck, W. Inflammation, demyelination, and degeneration - recent insights from ms pathology. Biochimica et biophysica acta 2011, 1812, 275-282. 407. Espitia Pinzon, N.; Stroo, E.; t Hart, B.A.; Bol, J.G.; Drukarch, B.; Bauer, J.; van Dam, A.M. Tissue transglutaminase in marmoset experimental multiple sclerosis: Discrepancy between white and grey matter. PloS one 2014, 9, e100574. 408. van Strien, M.E.; de Vries, H.E.; Chrobok, N.L.; Bol, J.; Breve, J.J.P.; van der Pol, S.M.P.; Kooij, G.; van Buul, J.D.; Karpuj, M.; Steinman, L., et al. Tissue transglutaminase contributes to experimental multiple sclerosis pathogenesis and clinical outcome by promoting macrophage migration. Brain, behavior, and

152

immunity 2015, 50, 141-154. 409. Chrobok, N.L.; Bol, J.; Jongenelen, C.A.; Breve, J.J.P.; El Alaoui, S.; Wilhelmus, M.M.M.; Drukarch, B.; van Dam, A.M. Characterization of transglutaminase 2 activity inhibitors in monocytes in vitro and their effect in a mouse model for multiple sclerosis. PloS one 2018, 13, e0196433. 410. Chrobok, N.L.; Jaouen, A.; Fenrich, K.K.; Bol, J.G.; Wilhelmus, M.M.; Drukarch, B.; Debarbieux, F.; van Dam, A.M. Monocyte behaviour and tissue transglutaminase expression during experimental autoimmune encephalomyelitis in transgenic cx3cr1(gfp/gfp) mice. Amino acids 2017, 49, 643-658. 411. Giera, S.; Luo, R.; Ying, Y.; Ackerman, S.D.; Jeong, S.J.; Stoveken, H.M.; Folts, C.J.; Welsh, C.A.; Tall, G.G.; Stevens, B., et al. Microglial transglutaminase-2 drives myelination and myelin repair via gpr56/adgrg1 in oligodendrocyte precursor cells. eLife 2018, 7. 412. Hwang, I.K.; Yoo, K.Y.; Yi, S.S.; Kim, I.Y.; Hwang, H.S.; Lee, K.Y.; Choi, S.M.; Lee, I.S.; Yoon, Y.S.; Kim, S.Y., et al. Expression of tissue-type transglutaminase (ttg) and the effect of ttg inhibitor on the hippocampal ca1 region after transient ischemia in gerbils. Brain research 2009, 1263, 134-142. 413. Oono, M.; Okado-Matsumoto, A.; Shodai, A.; Ido, A.; Ohta, Y.; Abe, K.; Ayaki, T.; Ito, H.; Takahashi, R.; Taniguchi, N., et al. Transglutaminase 2 accelerates neuroinflammation in amyotrophic lateral sclerosis through interaction with misfolded superoxide dismutase 1. Journal of neurochemistry 2014, 128, 403-418. 414. Curro, M.; Gangemi, C.; Gugliandolo, A.; Risitano, R.; Ferlazzo, N.; Ientile, R.; Caccamo, D. Transglutaminase 2 is involved in homocysteine-induced activation of human thp-1 monocytes. Free radical research 2015, 49, 299-308. 415. Crockett, J.C.; Rogers, M.J.; Coxon, F.P.; Hocking, L.J.; Helfrich, M.H. Bone remodelling at a glance. Journal of cell science 2011, 124, 991-998. 416. Jones, D.H.; Kong, Y.Y.; Penninger, J.M. Role of rankl and rank in bone loss and arthritis. Annals of the rheumatic diseases 2002, 61 Suppl 2, ii32-39. 417. Matarese, G.; Curro, M.; Isola, G.; Caccamo, D.; Vecchio, M.; Giunta, M.L.; Ramaglia, L.; Cordasco, G.; Williams, R.C.; Ientile, R. Transglutaminase 2 up-regulation is associated with rankl/opg pathway in cultured hpdl cells and thp-1-differentiated macrophages. Amino acids 2015, 47, 2447-2455. 418. Tanaka, H.; Shinki, T.; Takito, J.; Jin, C.H.; Suda, T. Transglutaminase is involved in the fusion of mouse alveolar macrophages induced by 1 alpha, 25-dihydroxyvitamin d3. Experimental cell research 1991, 192, 165-172. 419. Plugis, N.M.; Palanski, B.A.; Weng, C.H.; Albertelli, M.; Khosla, C. Thioredoxin-1 selectively activates transglutaminase 2 in the extracellular matrix of the small intestine: Implications for celiac disease. The Journal of biological chemistry 2017, 292, 2000-2008. 420. Gan, H.; Lee, J.; Ren, F.; Chen, M.; Kornfeld, H.; Remold, H.G. Mycobacterium tuberculosis blocks crosslinking of annexin-1 and apoptotic envelope formation on infected macrophages to maintain virulence. Nature immunology 2008, 9, 1189-1197. 421. Palucci, I.; Matic, I.; Falasca, L.; Minerva, M.; Maulucci, G.; De Spirito, M.; Petruccioli, E.; Goletti, D.; Rossin, F.; Piacentini, M., et al. Transglutaminase type 2 plays a key role in the pathogenesis of mycobacterium tuberculosis infection. Journal of internal medicine 2018, 283, 303-313. 422. Jeitner, T.M.; Pinto, J.T.; Cooper, A.J.L. Cystamine and cysteamine as inhibitors of transglutaminase activity in vivo. Bioscience reports 2018, 38. 423. Shrestha, C.L.; Assani, K.D.; Rinehardt, H.; Albastroiu, F.; Zhang, S.; Shell, R.; Amer, A.O.; Schlesinger, L.S.; Kopp, B.T. Cysteamine-mediated clearance of antibiotic-resistant pathogens in human cystic fibrosis macrophages. PloS one 2017, 12, e0186169.

153

424. Ferrari, E.; Monzani, R.; Villella, V.R.; Esposito, S.; Saluzzo, F.; Rossin, F.; D'Eletto, M.; Tosco, A.; De Gregorio, F.; Izzo, V., et al. Cysteamine re-establishes the clearance of pseudomonas aeruginosa by macrophages bearing the cystic fibrosis-relevant f508del-cftr mutation. Cell death & disease 2017, 8, e2544. 425. Hsu, T.C.; Chiang, S.Y.; Huang, C.Y.; Tsay, G.J.; Yang, C.W.; Huang, C.N.; Tzang, B.S. Beneficial effects of treatment with transglutaminase inhibitor cystamine on macrophage response in nzb/w f1 mice. Experimental biology and medicine (Maywood, N.J.) 2007, 232, 195-203. 426. Kim, S.Y. New insights into development of transglutaminase 2 inhibitors as pharmaceutical lead compounds. Medical sciences (Basel, Switzerland) 2018, 6. 427. Wodtke, R.; Hauser, C.; Ruiz-Gomez, G.; Jackel, E.; Bauer, D.; Lohse, M.; Wong, A.; Pufe, J.; Ludwig, F.A.; Fischer, S., et al. N(epsilon)-acryloyllysine piperazides as irreversible inhibitors of transglutaminase 2: Synthesis, structure-activity relationships, and pharmacokinetic profiling. Journal of medicinal chemistry 2018, 61, 4528-4560. 428. Song, M.; Hwang, H.; Im, C.Y.; Kim, S.Y. Recent progress in the development of transglutaminase 2 (tgase2) inhibitors. Journal of medicinal chemistry 2017, 60, 554-567. 429. Teitelbaum, S.L. Bone resorption by osteoclasts. Science (New York, N.Y.) 2000, 289, 1504-1508. 430. Ross, F.P.; Teitelbaum, S.L. Alphavbeta3 and macrophage colony-stimulating factor: Partners in osteoclast biology. Immunological reviews 2005, 208, 88-105. 431. Matsumoto, M.; Kogawa, M.; Wada, S.; Takayanagi, H.; Tsujimoto, M.; Katayama, S.; Hisatake, K.; Nogi, Y. Essential role of p38 mitogen-activated protein kinase in cathepsin k gene expression during osteoclastogenesis through association of nfatc1 and pu.1. The Journal of biological chemistry 2004, 279, 45969-45979. 432. Sundaram, K.; Nishimura, R.; Senn, J.; Youssef, R.F.; London, S.D.; Reddy, S.V. Rank ligand signaling modulates the matrix metalloproteinase-9 gene expression during osteoclast differentiation. Experimental cell research 2007, 313, 168-178. 433. Zhang, C.; Dou, C.E.; Xu, J.; Dong, S. Dc-stamp, the key fusion-mediating molecule in osteoclastogenesis. Journal of cellular physiology 2014, 229, 1330-1335. 434. Linder, S.; Aepfelbacher, M. Podosomes: Adhesion hot-spots of invasive cells. Trends in cell biology 2003, 13, 376-385. 435. Georgess, D.; Machuca-Gayet, I.; Blangy, A.; Jurdic, P. Podosome organization drives osteoclast- mediated bone resorption. Cell adhesion & migration 2014, 8, 191-204. 436. Hodrea, J.; Demeny, M.A.; Majai, G.; Sarang, Z.; Korponay-Szabo, I.R.; Fesus, L. Transglutaminase 2 is expressed and active on the surface of human monocyte-derived dendritic cells and macrophages. Immunology letters 2010, 130, 74-81. 437. Cui, C.; Kaartinen, M.T. Serotonin (5-ht) inhibits factor xiii-a-mediated plasma fibronectin matrix assembly and crosslinking in osteoblast cultures via direct competition with transamidation. Bone 2015, 72, 43-52. 438. Heath, D.J.; Downes, S.; Verderio, E.; Griffin, M. Characterization of tissue transglutaminase in human osteoblast-like cells. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2001, 16, 1477-1485. 439. Kaartinen, M.T.; El-Maadawy, S.; Rasanen, N.H.; McKee, M.D. Tissue transglutaminase and its substrates in bone. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research 2002, 17, 2161-2173. 440. Nakano, Y.; Al-Jallad, H.F.; Mousa, A.; Kaartinen, M.T. Expression and localization of plasma

154

transglutaminase factor xiiia in bone. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 2007, 55, 675-685. 441. Wang, Z.; Telci, D.; Griffin, M. Importance of syndecan-4 and syndecan -2 in osteoblast cell adhesion and survival mediated by a tissue transglutaminase-fibronectin complex. Experimental cell research 2011, 317, 367-381. 442. Tiedemann, K.; Boraschi-Diaz, I.; Rajakumar, I.; Kaur, J.; Roughley, P.; Reinhardt, D.P.; Komarova, S.V. Fibrillin-1 directly regulates osteoclast formation and function by a dual mechanism. Journal of cell science 2013, 126, 4187-4194. 443. Li, Z.; Li, C.; Zhou, Y.; Chen, W.; Luo, G.; Zhang, Z.; Wang, H.; Zhang, Y.; Xu, D.; Sheng, P. Advanced glycation end products biphasically modulate bone resorption in osteoclast-like cells. American journal of physiology. Endocrinology and metabolism 2016, 310, E355-366. 444. Kartner, N.; Yao, Y.; Li, K.; Crasto, G.J.; Datti, A.; Manolson, M.F. Inhibition of osteoclast bone resorption by disrupting vacuolar h+-atpase a3-b2 subunit interaction. The Journal of biological chemistry 2010, 285, 37476-37490. 445. Hulkower, K.I.; Herber, R.L. Cell migration and invasion assays as tools for drug discovery. Pharmaceutics 2011, 3, 107-124. 446. Wilhelmus, M.M.; de Jager, M.; Smit, A.B.; van der Loo, R.J.; Drukarch, B. Catalytically active tissue transglutaminase colocalises with abeta pathology in alzheimer's disease mouse models. Scientific reports 2016, 6, 20569. 447. van den Akker, J.; VanBavel, E.; van Geel, R.; Matlung, H.L.; Guvenc Tuna, B.; Janssen, G.M.; van Veelen, P.A.; Boelens, W.C.; De Mey, J.G.; Bakker, E.N. The redox state of transglutaminase 2 controls arterial remodeling. PloS one 2011, 6, e23067. 448. Aleman, M.M.; Byrnes, J.R.; Wang, J.G.; Tran, R.; Lam, W.A.; Di Paola, J.; Mackman, N.; Degen, J.L.; Flick, M.J.; Wolberg, A.S. Factor xiii activity mediates red blood cell retention in venous thrombi. The Journal of clinical investigation 2014, 124, 3590-3600. 449. Monteagudo, A.; Ji, C.; Akbar, A.; Keillor, J.W.; Johnson, G.V. Inhibition or ablation of transglutaminase 2 impairs astrocyte migration. Biochemical and biophysical research communications 2017, 482, 942- 947. 450. Cordella, M.; Tabolacci, C.; Rossi, S.; Senatore, C.; Facchiano, A.M.; D'Arcangelo, D.; Facchiano, A.; Facchiano, F. Transglutaminase type 2 affects cell migration through post-translational modification of platelet-derived growth factor-bb. Amino acids 2017, 49, 473-481. 451. Bernard, S.; Herzel, H. Why do cells cycle with a 24 hour period? Genome informatics. International Conference on Genome Informatics 2006, 17, 72-79. 452. Wang, S.; Cui, C.; Hitomi, K.; Kaartinen, M.T. Detyrosinated glu-tubulin is a substrate for cellular factor xiiia transglutaminase in differentiating osteoblasts. Amino acids 2014, 46, 1513-1526. 453. Kogawa, M.; Findlay, D.M.; Anderson, P.H.; Ormsby, R.; Vincent, C.; Morris, H.A.; Atkins, G.J. Osteoclastic metabolism of 25(oh)-vitamin d3: A potential mechanism for optimization of bone resorption. Endocrinology 2010, 151, 4613-4625. 454. Hoshino, A.; Ueha, S.; Hanada, S.; Imai, T.; Ito, M.; Yamamoto, K.; Matsushima, K.; Yamaguchi, A.; Iimura, T. Roles of chemokine receptor cx3cr1 in maintaining murine bone homeostasis through the regulation of both osteoblasts and osteoclasts. Journal of cell science 2013, 126, 1032-1045. 455. Kopesky, P.; Tiedemann, K.; Alkekhia, D.; Zechner, C.; Millard, B.; Schoeberl, B.; Komarova, S.V. Autocrine signaling is a key regulatory element during osteoclastogenesis. Biology open 2014, 3, 767- 776.

155

456. Siegbahn, A.; Hammacher, A.; Westermark, B.; Heldin, C.H. Differential effects of the various isoforms of platelet-derived growth factor on chemotaxis of fibroblasts, monocytes, and granulocytes. The Journal of clinical investigation 1990, 85, 916-920. 457. Cho, J.S.; Fang, T.C.; Reynolds, T.L.; Sofia, D.J.; Hamann, S.; Burkly, L.C. Pdgf-bb promotes type i ifn- dependent vascular alterations and monocyte recruitment in a model of dermal fibrosis. PloS one 2016, 11, e0162758. 458. Bethel-Brown, C.; Yao, H.; Hu, G.; Buch, S. Platelet-derived growth factor (pdgf)-bb-mediated induction of monocyte chemoattractant protein 1 in human astrocytes: Implications for hiv-associated neuroinflammation. Journal of neuroinflammation 2012, 9, 262. 459. Mizoguchi, F.; Murakami, Y.; Saito, T.; Miyasaka, N.; Kohsaka, H. Mir-31 controls osteoclast formation and bone resorption by targeting rhoa. Arthritis research & therapy 2013, 15, R102. 460. Janiak, A.; Zemskov, E.A.; Belkin, A.M. Cell surface transglutaminase promotes rhoa activation via integrin clustering and suppression of the src-p190rhogap signaling pathway. Molecular biology of the cell 2006, 17, 1606-1619. 461. Guilluy, C.; Rolli-Derkinderen, M.; Tharaux, P.L.; Melino, G.; Pacaud, P.; Loirand, G. Transglutaminase- dependent rhoa activation and depletion by serotonin in vascular smooth muscle cells. The Journal of biological chemistry 2007, 282, 2918-2928. 462. Guilluy, C.; Eddahibi, S.; Agard, C.; Guignabert, C.; Izikki, M.; Tu, L.; Savale, L.; Humbert, M.; Fadel, E.; Adnot, S., et al. Rhoa and rho kinase activation in human pulmonary hypertension: Role of 5-ht signaling. American journal of respiratory and critical care medicine 2009, 179, 1151-1158. 463. Yi, S.J.; Groffen, J.; Heisterkamp, N. Transglutaminase 2 regulates the gtpase-activating activity of bcr. The Journal of biological chemistry 2009, 284, 35645-35651. 464. Kanchan, K.; Fuxreiter, M.; Fesus, L. Physiological, pathological, and structural implications of non- enzymatic protein-protein interactions of the multifunctional human transglutaminase 2. Cellular and molecular life sciences : CMLS 2015, 72, 3009-3035. 465. Soysa, N.S.; Alles, N.; Aoki, K.; Ohya, K. Osteoclast formation and differentiation: An overview. Journal of medical and dental sciences 2012, 59, 65-74. 466. Hamilton, T.A.; Zhao, C.; Pavicic, P.G., Jr.; Datta, S. Myeloid colony-stimulating factors as regulators of macrophage polarization. Frontiers in immunology 2014, 5, 554. 467. Yu, T.; Wang, W.; Nassiri, S.; Kwan, T.; Dang, C.; Liu, W.; Spiller, K.L. Temporal and spatial distribution of macrophage phenotype markers in the foreign body response to glutaraldehyde-crosslinked gelatin hydrogels. Journal of biomaterials science. Polymer edition 2016, 27, 721-742. 468. Huang, R.; Wang, X.; Zhou, Y.; Xiao, Y. Rankl-induced m1 macrophages are involved in bone formation. Bone research 2017, 5, 17019. 469. Zheng, H.; Yu, X.; Collin-Osdoby, P.; Osdoby, P. Rankl stimulates inducible nitric-oxide synthase expression and nitric oxide production in developing osteoclasts. An autocrine negative feedback mechanism triggered by rankl-induced interferon-beta via nf-kappab that restrains osteoclastogenesis and bone resorption. The Journal of biological chemistry 2006, 281, 15809-15820. 470. Basso, M.; Berlin, J.; Xia, L.; Sleiman, S.F.; Ko, B.; Haskew-Layton, R.; Kim, E.; Antonyak, M.A.; Cerione, R.A.; Iismaa, S.E., et al. Transglutaminase inhibition protects against oxidative stress-induced neuronal death downstream of pathological erk activation. The Journal of neuroscience : the official journal of the Society for Neuroscience 2012, 32, 6561-6569. 471. Lai, T.S.; Liu, Y.; Tucker, T.; Daniel, K.R.; Sane, D.C.; Toone, E.; Burke, J.R.; Strittmatter, W.J.; Greenberg, C.S. Identification of chemical inhibitors to human tissue transglutaminase by screening

156

existing drug libraries. Chemistry & biology 2008, 15, 969-978. 472. Stamnaes, J.; Iversen, R.; du Pre, M.F.; Chen, X.; Sollid, L.M. Enhanced b-cell receptor recognition of the autoantigen transglutaminase 2 by efficient catalytic self-multimerization. PloS one 2015, 10, e0134922. 473. Sun, H.; Kaartinen, M.T. Transglutaminases in monocytes and macrophages. Medical sciences (Basel, Switzerland) 2018, 6. 474. Hwang, S.Y.; Putney, J.W., Jr. Calcium signaling in osteoclasts. Biochimica et biophysica acta 2011, 1813, 979-983. 475. Kim, S.J.; Kim, K.H.; Ahn, E.R.; Yoo, B.C.; Kim, S.Y. Depletion of cathepsin d by transglutaminase 2 through protein cross-linking promotes cell survival. Amino acids 2013, 44, 73-80.

157