Nemosis As a Model for Fibroblast Inflammatory Activation

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Helsingin yliopiston digitaalinen arkisto

Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation

Anna Enzerink

Department of Virology Haartman Institute & Division of Biochemistry Department of Biosciences Faculty of Biological and Environmental Sciences

University of Helsinki Finland

Helsinki Graduate School in Biotechnology and Molecular Biology

Academic dissertation

To be presented for public examination, with the permission of the Faculty of Biological and Environmental Sciences of the University of Helsinki, in Lecture Hall 2, Haartman Institute, Haartmaninkatu 3, Helsinki, on August 13th, at 12 noon. Helsinki 2010 Supervised by Professor emeritus Antti Vaheri Department of Virology Haartman Institute University of Helsinki Finland

Reviewed by Professor Pirjo Laakkonen Professor Hannu Larjava Molecular Cancer Biology Program Department of Oral Biological Biomedicum Helsinki and Medical Sciences University of Helsinki Faculty of Dentistry University of British Columbia, Vancouver Department of Biotechnology Canada and Molecular Biology A. I. Virtanen Institute for Molecular Sciences University of Eastern Finland, Kuopio Finland

Opponent Custos Professor Christopher Buckley Professor Carl Gahmberg Institute of Biomedical Research Division of Biochemistry School of Immunity and Infection Department of Biosciences University of Birmingham Medical School Faculty of Biological and United Kingdom Environmental Sciences University of Helsinki Finland

Lay-out Arno Enzerink

Language revision Mary Metzler, PhD / University of Helsinki Language Centre

Cover illustration Catherine Nobes & Alan Hall / Wellcome Images Activated rat embryo fibroblast showing actin stress fibers (red) and focal adhesions (vinculin: green/yellow)

Yliopistopaino ISBN 978-952-92-7603-5 (paperback) ISBN 978-952-10-6388-6 (PDF) http://ethesis.helsinki.fi Helsinki 2010

- W. Shakespeare (Hamlet) Contents

Abbreviations 6

List of original publications 8

Abstract 9

Review of the literature 11 Early steps to acute inflammation: the interplay between leukocytes and endothelium 11 Activation of the endothelial surface leads to enhanced adhesion of leukocytes 12 Decreased barrier function of the endothelium aids leukocyte extravasation 14 Chemokines promote leukocyte chemotaxis 15 Fibroblasts are modulators of the inflammation reaction 17 Fibroblasts are tissue-resident sentinel cells 18 Stromal fibroblasts in inflammation: the myofibroblast phenotype 19 Later stages in tissue repair: neovascularization of the stroma 20 Dynamics of the endothelium during angiogenesis 21 Intercellular signaling molecules regulating angiogenesis 22 Role of stromal fibroblasts in angiogenesis 25 Nemosis is a model for the stromal activation process of fibroblasts 25 Early steps in nemosis of fibroblasts: interactions with the extracellular matrix 28 Influence of nemotic fibroblasts on epithelial cells 29 Influence of nemotic fibroblasts on malignant cells 29

Aims of the study 31

Materials and methods 33 Cell isolation, culture and initiation of nemosis in fibroblasts (I, II, III) 33 Microarray experiments and analysis (I) 34 Quantification of growth factors and chemokines in conditioned fibroblast medium (I, III) 35 Chemotaxis assay (I) 35 Determination of IκB levels and NF-κB activity (I) 36 Analysis of tight and adherens junctions on endothelial monolayers (II) 36 Static leukocyte adhesion assay (II) 37 Scratch-wound repair assay and live-cell imaging (III) 37 Endothelial sprout formation assay (III) 38 Proliferation assay (III) 38

Results and discussion 39 Nemotic fibroblasts exhibit a proinflammatory response (I, II, IV) 39 Production of chemotactic cytokines and attraction of leukocytes 39 Activation of endothelium and modulation of its permeability 42 Fibroblast nemosis has angiogenesis-promoting properties (III) 45 Nemotic fibroblasts promote angiogenic properties of endothelial cells 45 Other angiogenesis-promoting factors involved in nemosis 47 The dark side: inflammation-related damage and tumorigenesis 48

Concluding remarks 53

Acknowledgements 56

References 60

Original publications 69 Abbreviations

α-SMA α-smooth-muscle actin BAEC bovine aortic endothelial cell BSA bovine serum albumin CAF cancer-associated fibroblast CCL chemokine ligand CCR chemokine receptor COX cyclooxygenase DMEM Dulbecco’s modified Eagle medium ECL enhanced chemiluminescence ECM extracellular matrix EDTA ethylenediaminetetraacetic acid ELISA enzyme-linked immunosorbent assay FBS fetal bovine serum FGF fibroblast growth factor FITC fluorescein isothiocyanate GFP green fluorescent protein GRO growth-related oncogene HAL human adult lung HES human embryonal skin HFSF human foreskin fibroblast HGF hepatocyte growth factor HMEC-d human microvascular endothelial cell - dermal HRP horseradish peroxidase HTAB hexa-decyl-trimethyl-ammonium bromide HUVEC human umbilical vein endothelial cell ICAM intercellular adhesion molecule IL interleukin IκB inhibitor of NF-κB MAb monoclonal antibody MET mesenchymal-epithelial transition factor MCP monocyte chemotactic protein MIP macrophage inflammatory protein MMP matrix metalloproteinase MT-MMP membrane type matrix metalloproteinase NF-κB nuclear factor-κB PBS phosphate-buffered saline PCNA proliferating cell nuclear antigen PG (D, E, F) prostaglandin (D, E, F) RANTES regulated upon activation, normal T cell expressed and secreted RPMI Roswell Park Memorial Institute (medium) SDF stromal-derived factor SDS-PAGE sodium dodecyl sulphate polyacrylamide gel electrophoresis SEM standard error of the mean TNF-α tumor necrosis factor-α VCAM vascular cell adhesion molecule VE-cadherin vascular-endothelial cadherin VEGF vascular endothelial growth factor VEGFR vascular endothelial growth factor receptor ZO zonula occludens List of original publications

This thesis is based on the following original research articles and reviews, which are referred to in the text by their assigned Roman numerals.

I Enzerink A, Salmenperä P, Kankuri E, Vaheri A. Clustering of fibroblasts induces proinflammatory chemokine secretion promoting leukocyte migration. Molecular Immunology. 2009, 46:1787-1795.

II Enzerink A, Vaheri A. Fibroblast nemosis promotes proinflammatory properties of endothelial cells by modulating cell-cell junctions and leukocytes by augmenting adhesion. Manuscript.

III Enzerink A, Rantanen V, Vaheri A. Fibroblast nemosis induces angiogenic responses of endothelial cells. Experimental Cell Research. 2010, 316:826-835.

IV Vaheri A, Enzerink A, Räsänen K, Salmenperä P. Nemosis, a novel way of fibroblast activation, in inflammation and cancer. Experimental Cell Research.

2009, 315:1633-1638. *Equal contribution between all authors

Wound healing is a complex process that requires an interplay between several cell types. Classically, fibroblasts have been viewed as producers of extracellular matrix, but more recently they have been recognized as orchestrators of the healing response, promoting and directing inflammation and neovascularization processes. Compared to those from healthy tissue, inflammation-associated fibroblasts display a dramatically altered phenotype and have been described as sentinel cells, able to switch to an immunoregulatory profile on cue. However, the activation mechanism still remains largely uncharacterized.

Nemosis is a model for stromal fibroblast activation. When normal human primary fibroblasts are deprived of growth support they cluster, forming multicellu- lar spheroids. Clustering results in upregulation of proinflammatory markers such as cyclooxygenase-2 and secretion of prostaglandins, proteinases, cytokines, and growth factors. Fibroblasts in nemosis induce wound healing and tumorigenic responses in many cell types found in inflammatory and tumor microenvironments.

This study investigated the effect of nemotic fibroblasts on two components of the vascular system, leukocytes and endothelium, and characterized the inflammation-promoting responses that arose in these cell types. Fibroblasts in nemosis were found to secrete an array of chemotactic cytokines and attract leukocytes, as well as promote their adhesion to the endothelium. Nuclear factor-κB, the master regulator of many inflammatory responses, is activated in nemotic fibroblasts. Nemotic fibroblasts are known to produce large amounts of hepatocyte growth factor, a motogenic and angiogenic factor. Also, as shown in this study, they produce vascular endothelial growth factor. These two factors induced migratory and sprouting responses in endothelial cells, both required for neovascularization. Nemotic fibroblasts also caused a decrease in the expression of adherens and tight junction components on the surface of endothelial cells. The results allow the conclusion that fibroblasts in nemosis share many similarities with inflammation-associated fibroblasts. Both inflammation and stromal fibroblasts are known to be involved in tumorigenesis and tumor progression. Ne- mosis may be viewed as a model for stromal fibroblast activation, or it may cor- relate with cell-cell interactions between adjacent fibroblastsin vivo. Nevertheless, due to nemosis-derived production of proinflammatory cytokines and growth factors, fibroblast nemosis may have therapeutic potential as an inducer of controlled tissue repair. Knowledge of stromal fibroblast activation gained through studies of nemosis could provide new strategies to control unwanted inflammation and tumor progression. Review of the literature

The importance of wound care was known by healers in Ancient Egypt, according to the first records, dating from 1400 BC. However, the concept of hemostasis was not understood, and bleeding was most likely only poorly controlled. The first written record regarding the importance of hemorrhage control dates back to 1500 BC, and this papyrus from Egypt contains instructions on how to stop bleeding using heated instruments and fire. In Ancient Greece, bleeding was already controlled by mechanical means. From records that remain today it can be deduced that the Roman physician Cornelius Celsus was probably the first to mention the two hallmarks of wound care: treatment of hemorrhage and control of inflammation (Broughton et al., 2006). Sadly, most knowledge was lost or forgotten during the Middle Ages, and scientific interest in wound care, hemostasis, and inflammation was re-ignited again only approximately 500 years ago. Over the last century the links between cancer and inflammation have gradually become evident (reviewed in Balkwill and Mantovani, 2001). In addition, leukocyte-endothelial interactions (Cohnheim, 1867), angiogenesis (Lang, 2001), and the importance of fibroblasts in inflammation (Gabbiani et al., 1971; Smith et al., 1997) were discovered, ultimately resulting in the rapid accumulation of scientific information and care practices.

Early steps to acute inflammation: the interplay between leukocytes and endothelium

The process of wound repair is a tightly controlled process, which is required when tissue function is impaired by e.g. insult or infection (Li et al., 2007; Med- zhitov, 2008). Acute inflammation has evolved to act as a driving force during the repair process. Wound healing has partially overlapping phases with clearly defined functions, beginning with the hemostasis response and ending in the resolution of inflammation, and healed tissue. In the case of tissue injury, the first occurrence is often vessel disruption, which results in a hemostasis response: bleeding, platelet aggregation, and finally clot formation. The first steps in the inflammation response begin already with the proteolytic cascades required for clot formation, but cells

Review of the literature | 11 present at the injury site must recognize the damage in order to alert the inflammatory defense system.

Activation of the endothelial surface leads to enhanced adhesion of leukocytes

One important cell type in the first line of defense is the endothelial cell, which lines blood vessel walls. In healthy tissue, the endothelium forms a semipermeable membrane that allows the exchange of small molecules and the entrance of a low amount of patrolling macrophages. A crucial component of inflammation is the extravasation of inflammatory leukocytes such as polymorphonuclear neutrophils and monocytes, which invade the inflamed tissue and surrounding stroma where they take part in host defense and tissue repair (Figure 1) (Chavakis et al., 2009; Granger and Kubes, 1994; Muller, 2009). The activation of monocytes to macrophages is especially important for the inflammatory response, because an activated, tissue-resident macrophage not only clears tissue of pathogens and debris, but also takes part in controlling the inflammatory response by way of cytokine and growth factor secretion (Serbina et al., 2008). For these cells to recognize the

Healthy vasculature Inﬂammation

circulating activated leukocyte endothelium rolling ﬁrm adhesion (selectins) (integrins) extravasation (adherens- & tight junctions) endothelium basement membrane

ﬁbroblast connective tissue activated extravasated, smooth muscle cells ﬁbroblast activated leukocyte

Figure 1. Schematic model of leukocyte-endothelial interactions on the vessel wall in healthy and inflamed vasculature, and the influence of the underlying stromal fibroblasts.

12 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation site of extravasation, the endothelium must transmit distress signals, and thus be in an activated state. Endothelial cells are anchored to a basement membrane and require constant contact with the extracellular matrix (ECM) for proper function. The endothelial cell is adaptive and responsive to its microenvironment: it displays various phenotypes depending on the type of vasculature or the organ it resides in (Langenkamp and Molema, 2009; Pai and Ruoslahti, 2005). Activation of the endothelium causes the cells to shift phenotype or function (Zimmerman et al., 1999). The outcome is a dynamic process that is tailored to meet a specific need, related to the specific microenvironmental stimulus in question. Therefore, there is no single, specific characterization of the process of endothelial activation. Classical activating agents are thrombin, histamine, tumor necrosis factor-α (TNF-α), and various cytokines, all of which result in slightly different profiles of activation. The response varies temporally and the first response is rapid. Some activation markers, such as platelet-selectin, are released from intracellular stores and upregulated in a matter of minutes (Hattori et al., 1989; Zimmerman et al., 1999). The proinflammatory transcription factor nuclear factor κB (NF-κB) is also often activated within minutes (Li, 2002; Sen and Baltimore, 1986). Other markers, such as the cell-surface deposition of intercellular adhesion molecule-1 (ICAM-1), require de novo protein synthesis and thus a time-span of several hours for proper activation (Dustin et al., 1986). Such temporal differences are important in orchestrating a controlled, sequential inflammatory response. One purpose served by proinflammatory endothelial activation is to increase the adhesion of leukocytes, enabling their activation and subsequent extravasation into inflamed tissue (Chavakis et al., 2009; Granger and Kubes, 1994; Muller, 2009). The major site for extravasation is the postcapillary venule, where lower shear stress results in enhanced interactions with the vessel wall (Granger and Kubes, 1994). The adhesion process for leukocytes is initiated with a weak adhesive interaction that causes the cell to roll over the endothelium and is followed by a firm, high-adhesive sticking to the vessel wall. Rolling is mediated by selectins (leukocyte-, endothelial- and platelet-selectins) and their glycoprotein ligands, and results in a reduction in the cell’s flow speed, allowing further, firmer adhesive -in teractions to take place (Vestweber and Blanks, 1999). ICAM-1 on the endothelial cell, and its receptors on the leukocyte, the β2-integrins (CD11/CD18), are mol-

Review of the literature | 13 ecules that mediate firm adhesion. ICAM-1 is a glycoprotein widely expressed on the endothelial cell surface in an active form, and its expression can be induced by stimuli such as exposure to cytokines (Dustin et al., 1986; Lawson and Wolf, 2009;

Patarroyo et al., 1987). The β2-integrin group consists of various CD11 subunits bound to a common CD18 subunit (Gahmberg et al., 1997). When activated by the

contact with the endothelium, the β2-integrin undergoes a conformational change resulting in a higher affinity for its ligand. Adhesion can be further augmented by

membrane clustering of multiple active β2-integrin units (Gahmberg et al., 2009). As a result, the leukocyte adheres to the vessel wall, flattens, crawls to an appro- priate area, and attempts to transmigrate through the endothelium into inflamed tissue (Muller, 2009).

Decreased barrier function of the endothelium aids leukocyte extravasation

For successful extravasation, leukocytes need to be allowed access through a monolayer of endothelial cells. Leukocytes use two pathways to transmigrate through the endothelial layer. The most studied pathway is the paracellular pathway, where the leukocyte pushes through the endothelium between adjacent cells. Transmigration can also take place transcellularly, through the endothelial cell body, seemingly by utilizing vesicular vacuolar organelle structures (Feng et al., 1998; Garrido-Urbani et al., 2008), but this pathway seems to be less common (Carman and Springer, 2004). Extravasation requires the endothelium to increase its permeability around the site of injury (Aghajanian et al., 2008; Wallez and Huber, 2008). The physical barrier function of the endothelium, including control of the flow of ions, macromol- ecules, and blood cells, is regulated by tight junctions between endothelial cells (Bazzoni, 2006; Farquhar and Palade, 1963). The claudin family transmembrane proteins are the building blocks of tight junctions. They form strands by interacting both with each other on the same cell (heterophilic binding) as well as by binding two cells tightly together (homophilic binding) (Findley and Koval, 2009). Typically, each claudin-expressing tissue, such as epithelium and endothelium, harbors several types of claudin, and the expression profile is tissue-specific. The importance of the endothelial cell-specific claudin-5 in regulating endothelial permeability is well

14 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation illustrated in blood-brain-barrier models, where claudin-5 deficiency causes im- pairment of the blood-brain-barrier (Morita et al., 1999; Nitta et al., 2003). Clau- dins are associated with ZO-1 (zonula occludens-1) on the cytosolic side (Fanning and Anderson, 2009; Itoh et al., 1999). ZO-1 is a scaffolding protein that binds a number of tight junction proteins and orchestrates the assembly of tight junctions into functional strands at cell-cell junctions (Umeda et al., 2006). Endothelial tight junctions are spatially intermingled with adherens junctions, which control cell-cell adhesion of the monolayer, as well as with several signaling pathways regulating growth and differentiation (Dejana et al., 2008; Farquhar and Palade, 1963). There are several types of endothelial adherens junctions, and a major type contains the transmembrane protein vascular endothelial (VE) cadherin. VE-cadherin junctions are believed to play a major role in stabilizing endothelial cell contacts and thus vascular integrity and physiological functions (Corada et al., 1999; Lampugnani et al., 1992; Vestweber, 2008). VE-cadherin forms homophilic contacts on the cell surface, and adhesion is dependent on direct association with β-catenin on the intracellular side (Ozawa et al., 1989; Xu and Kimelman, 2007). Functioning as a scaffold protein, β-catenin links VE-cadherin to the actin cyto- skeleton. ZO-1 is also known to interact with adherens junction proteins during the formation of junctional complexes, and may directly regulate activity of mature adherens junctions (Fanning and Anderson, 2009). VE-cadherin and β-catenin act together as gatekeepers for leukocyte transmigration, and have been shown to be transiently removed from cell-cell junctions at the site of transmigration (Allport et al., 2000).

Chemokines promote leukocyte chemotaxis

Cytokines are signaling molecules that mediate the messages in a complicated cross-talk between all cell types involved in the inflammatory response. For an effective inflammatory host response to take place, leukocytes must be directed to the site of injury. Extravasated and tissue-resident leukocytes are attracted by an increasing gradient of chemotactic cytokines or chemokines (Table 1, see following page). There are two major groups of chemokines: the CXC chemokines and the CC chemokines, named after differing peptide motifs in the respective proteins (Thor-

Review of the literature | 15 pe et al., 2003). Chemokines both within and between groups share approximately 20-50% homology on the amino acid level (Strieter, 2002). Their targets are different: the CXC chemokines, e.g. CXCL1, 2, 3 (growth-related oncogene (GRO) -α, -β, -γ), and CXCL8, are classically viewed as granulocyte and eosinophil specific, binding to the G-protein-coupled receptors CXCR1-6 on the target cells. The CC chemokines, e.g. CCL2 (monocyte chemoattractant protein-1; MCP-1), CCL3 (macrophage inflammatory protein-1α; MIP-1α), and CCL5 (regulated upon activation, normal T cell expressed and secreted; RANTES), are mainly directed towards monocyte-macrophages and lymphocytes, and bind to the receptors CCR1-11 on the leukocyte (Rossi and Zlotnik, 2000; Sharma, 2010). Chemokines are secreted into the microenvironment, but they can also be presented on the surface of endothelial cells, in association with proteoglycans, causing rolling and adhering leukocytes to activate and prepare for battle (Rot and von Andrian, 2004; Tanaka et al., 1993).

Main Chemokine Alternative Main leukocyte name name sources targets Receptors CXCL1 GRO-a EC, FB, M N CXCR1,2 CXCL2 GRO-b EC, FB, M N CXCR2 CXCL3 GRO-g EC, FB, M N CXCR2 CXCL8 IL-8 EC, FB, KC, M, N MP, N CXCR1,2 CCL2 MCP-1 M, EC, SMC M CCR2 CCL3 MIP-1-a L, M, MP M CCR1,5 CCL4 MIP-1-b L, M, MP M CCR5 CCL5 RANTES FB, TL B, EO,M, TL CCR1,3,5

Table 1. Chemokines discussed in this study.

B (basophil), EC (endothelial cell), EO (eosinophil), FB (fibroblast), KC (keratinocyte), L (lymphocyte) M (monocyte) MP (macrophage), N (neutrophil), SMC (smooth muscle cell), TL (T- lymphocyte)

(Thorpe et al., 2003; COPE: Horst Ibelgaufts’ Cytokines & Cells Online Pathfinder Encyclopae- dia, www.copewithcytokines.org)

16 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation The expression of many chemokines is regulated by the transcription factor NF- κB, a key regulator of various pathogen-response and inflammation-related genes (Gilmore, 2006; Li, 2002; Sen and Baltimore, 1986). The NF-κB complex consists of two subunits, and these subunits are composed of combinations of p65 (RelA), p50/p105 (NF-κB1), p52/p100 (NF-κB2), c-Rel, or RelB. All proteins can form het- erodimers, and all except RelB can form homodimers. Activated NF-κB consists mainly of p65 together with either p50 or p52. The NF-κB complex remains inactive in the cytoplasm, bound to its endogenous inhibitor IκB (inhibitor of NF-κB). When activated upon various stimuli, either by the cell itself or by an external signal, IκB is ubiquitinated, which leads to release of NF-κB and rapid degradation of IκB in the proteasome. The liberated NF-κB translocates to the nucleus, where it binds to DNA and regulates the transcription of a plethora of target genes. Hence, NF-κB plays a pivotal role in triggering and maintaining inflammation, and it is often found to be constitutively active in inflammatory disorders (Hayden et al., 2006).

Fibroblasts are modulators of the inflammation reaction

It is still a challenge to define exactly what a fibroblast is. It is of mesenchymal or neural crest origin, and usually recognized morphologically as an adherent, flat, spindle-shaped cell with a flat, oval nucleus. When examined histologically, the fibroblast lacks the endothelial cell marker CD34, the epithelial cell marker cytok- eratin, and the hematopoietic cell marker CD45, but is often vimentin-positive. Yet, fibroblasts are by no means a single cell type, but display a heterogeneous phenotype that varies between organs and even within a single tissue. The fibroblast population in one organism is made up of various subsets of cells, each with distinct protein expression profiles and differing functions (Chang et al., 2002). Previously, it was thought that the sole function of a fibroblast was to produce and maintain the ECM, by which they are interconnected, producing components such as collagens and fibronectin. However, it has recently been appreciated that fibroblasts are also active initiators, controllers, and modulators of the inflammatory defense response (Baglole et al., 2006; Flavell et al., 2008). The traditional

Review of the literature | 17 view of vascular inflammatory reactions has been that of an inside-out response, beginning with endothelial activation, leukocyte extravasation, and an inflammatory response that spreads outwards from the vasculature. Recently, the opposite has been suggested: in the outside-in hypothesis, adventitial tissue, especially fibroblasts and leukocytes, are activated and influence vascular inflammation responses exogenously (Maiellaro and Taylor, 2007). Although further research is required to properly characterize this pathway, it presents an interesting perspective on the role of fibroblasts in triggering vascular inflammation responses.

Fibroblasts are tissue-resident sentinel cells

Unstimulated fibroblasts in healthy tissues produce a basal level of chemotactic cytokines such as CXCL8, CXCL12, CCL3, and CCL2, and the pattern of chemokines expressed varies with tissue type (Brouty-Boyé et al., 2000; Lukacs et al., 1994). Fibroblasts derived from inflammatory tissue have an altered chemokine profile (Hogaboam et al., 1998). Macrophages co-cultured with fibroblasts induce contact-dependent expression of cytokines, especially CCL3 (Hogaboam et al., 1998). Therefore, it has been suggested that fibroblasts play a role as sentinel cells, capable of switching to a phenotype that actively promotes an inflammatory response by producing cytokines, growth factors, and proteases under conditions such as tissue repair after injury, fibrosis, pathological organ remodeling, and cancer (Orimo and Weinberg, 2006; Smith et al., 1997). Fibroblasts are also capable of dysregulating inflammation, as displayed by mice injected with fibroblasts lacking the inflammatory regulator RelB. The result is a large local accumulation of inflammatory cells. RelB-/- fibroblasts were also shown to produce enhanced numbers of proinflammatory chemokines (Xia et al., 1997). This study shows the importance of controlling the fibroblast-derived inflammatory response. The role of fibroblasts as tissue-resident sentinel cells was first proposed by Smith et al. (Smith et al., 1997). They support this theory with four observations: firstly, fibroblasts form a heterogeneous population, and secondly, they display di- verse phenotypes in different anatomical regions. Thirdly, fibroblasts are capable of activation, upon which chemokines are produced and released, and fourthly, such activated fibroblasts are capable of regulating and maintaining leukocyte infiltra-

18 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation tion in damaged tissue. Thus, fibroblasts form an extended inflammatory defense system and act as an early warning system by secreting chemokines and other proinflammatory molecules that act in a paracrine fashion to alert the surrounding cells and tissues of immediate danger (Smith et al., 1997; Xia et al., 1997). Fibroblasts in many tissues express cyclooxygenase-1 (COX-1) and display basal levels of some of its end products, such as prostaglandin E2 (PGE2) (Wang et al., 1996). Several types of fibroblasts, including those from lung, synovia, and orbita are known to also express the inducible COX-2 isoform upon stimulation (Crofford et al., 1994; Wang et al., 1996; Zhang et al., 1998). Like many cytokines, COX-2 is also upregulated by NF-κB (Newton et al., 1997), which enhances the gatekeeper role of the latter in orchestrating the inflammatory response.

Stromal fibroblasts in inflammation: the myofibroblast phenotype

Gabbiani et al. discovered a type of fibroblast with a contractile function in healing wounds (Gabbiani et al., 1971; Gabbiani et al., 1978). This kind of fibroblast was further characterized by Sappino et al., who found similar, smooth muscle-like stromal cells expressing α-smooth muscle actin (α-SMA) in malignant breast tissue (Sappino et al., 1988). Due to their muscle-like characteristics they were named “myofibroblasts”. In contrast to fibroblasts in normal adult tissues, myofibroblasts in healing wounds make direct cell-cell contacts, such as through adherens and gap junctions (Ehrlich and Diez, 2003; Gabbiani et al., 1978; Hinz and Gabbiani, 2003). Typically, myofibroblasts also display a secretory phenotype: they produce chemokines and prostaglandins, and express both COX-1 and COX-2 isoforms (Powell et al., 1999). Differentiation towards the myofibroblast phenotype can be stimulated by various agents such as transforming growth factor-β, the ED-A splice variant of fibronectin, and changes in mechanical and tensile stress in the microenvironment (Hinz, 2007a; Serini et al., 1998; Tomasek et al., 2002). Fibroblast-to-myofibroblast differentiation in the wound-healing environment is a generally acknowledged phenomenon. While fibroblasts in various pathological tissue environments display some classical myofibroblast markers (e.g. α-SMA), they do not always express other such markers (e.g. fibronexi, gap junctions, prominent rough endoplasmic

Review of the literature | 19 reticulum). Also, other cell types have been shown to develop a spindle-shaped, α-SMA phenotype (Eyden, 2001). Due to these complications, the exact myofibroblast phenotype has not yet been clearly defined (Eyden, 2008). Currently, the precise origins of the myofibroblast are also not known. It is assumed that a common pathway for myofibroblast differentiation is from local, tissue-resident fibroblasts, but myofibroblasts may also arise from epithelial cells via epithelial-mesenchymal transition, or differentiate from pericytes, smooth-muscle cells, or circulating fi- brocytes (Gabbiani, 2004; McAnulty, 2007). However, fibroblasts displaying myofibroblast characteristics are abundant in inflammatory, fibrous, and malignant stroma (Hinz et al., 2007b; Powell et al., 1999). This highlights the importance of regarding fibroblasts as a heterogeneous group of cells with possibly slightly vary- ing functions within the stroma.

Later stages in tissue repair: neovascularization of the stroma

The process of tissue repair requires intense proliferation of several cell types and construction of a temporary, stromal microenvironment to support the healing process. Activated fibroblasts gather in and around the injured site, and together with inflammatory blood cells they form granulation tissue, which serves as a scaffold for the restoration of functional, healthy tissue. Traditionally, it has been thought that wound re-epithelialization and keratinocyte migration require the support of granulation tissue (Gurtner et al., 2008), but in many cases it seems that re-epithelialization is in fact completed before granulation tissue forms, and that granulation tissue is more important in later stages of wound repair such as dermal closure and skin remodeling (Braiman-Wiksman et al., 2007; Larjava et al., 1993). Healing and remodeling in turn place demands on the nutrition and oxygen supply in the wound bed. Local ischemia is common, especially in ruptured tissue with damaged blood vessels. These shortages are corrected by angiogenesis, resulting in neovascularization of the inflammatory microenvironment. Angiogenesis in tissue repair is activated by the innate inflammatory defense system and proceeds as a tightly controlled process, where endothelial cells, smooth muscle cells, and pericytes are guided by the microenvironment to sprout, migrate, and form new capillaries (Figure 2) (Adams and Alitalo, 2007; Eming et al., 2007).

20 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation activated inﬂammatory cell vascular lumen connective tissue

pericyte tip cell

activated vessel wall ﬁbroblast

stalk cells newly formed lumen

Figure 2. Schematic model of endothelial sprouting and neovascularization in response to signals from stromal fibroblasts and inflammatory cells.

Dynamics of the endothelium during angiogenesis

The purpose of larger blood vessels is to transport oxygen, nutrients, and waste products to or from tissues. The actual exchange between vasculature and tissue happens in the capillary network. These tiny blood vessels play a crucial role in providing oxygen and nutrients to surrounding tissue, and they also possess a powerful capability to sprout and branch on demand. Proangiogenic stimulation will result in altered endothelial cell permeability and loss of microvascular integrity, which is required for the formation of new vessels (Wallez and Huber, 2008). Sprout- ing, the first constructive step in neovascularization, requires dramatic alterations in the endothelial cells involved, including the ability to invade and migrate, as well as to degrade surrounding ECM (Adams and Alitalo, 2007). Upon angiogenic stimulation, a selection is required to define one tip cell to lead the way for a new sprout. The surrounding cells must stay in line and suppress their sprouting signaling to prevent serious disorder and formation of a leaking vessel. Recently, it was found that the tip cell expresses increased levels of Notch-ligand Delta-like-4,

Review of the literature | 21 whereas the following stalk cells have activated Notch receptor signaling. The signaling pathway and balance between these two molecules forms the communica- tion that keeps chaotic vessel formation at bay and provides stability and direction to the sprouting response (Hellstrom et al., 2007; Sainson et al., 2005). Vascular endothelial growth factor (VEGF) is the master regulator of the sprouting response, as it is both essential for the induction of tip cells, as well as for directing the sprout through a gradient (Gerhardt et al., 2003; Ruhrberg et al., 2002). Sprouting is followed by networking, where endothelial sprouts form connections with each other and pre-existing vessels. Lumen formation occurs simultaneously. Finally, newly formed vessels mature and are stabilized by the attachment of pericytes onto the outer surface. Controlled vascular permeability is not only important during leukocyte extravasation, but is also crucial during angiogenesis (Nyqvist et al., 2008). For successful endothelial migration and sprouting, cell-cell junctions must loosen up, but uncontrolled detachment of cell-cell junctional structures results in disintegrated and leaky vessels. Developmental models illustrate the importance of these factors in regulating vascular integrity and angiogenesis. VE-cadherin is important in embryonal angiogenesis, and VE-cadherin-deficiency or non-functionality is lethal in mice (Carmeliet et al., 1999). ZO-1 is also crucial for normal development: the null mutation is embryonally lethal, resulting in defective yolk-sac angiogenesis (Kat- suno et al., 2008). Angiogenic modulation of vascular integrity also plays a role in adulthood, especially in pathological conditions where permeability is chronically dysregulated, such as inflammatory diseases and cancer (Nagy et al., 2008).

Intercellular signaling molecules regulating angiogenesis

The first angiogenic factor was discovered in 1971 by Folkman et al., who isolated a soluble agent from neoplasms responsible for endothelial cell proliferation. This factor turned out to be VEGF, a potent driver of angiogenesis, which binds to the tyrosine kinase VEGFR2 on endothelial cells (Ferrara et al., 2003; Folkman et al., 1971). It is a homodimeric glycoprotein that comes in many splice isoforms, some of which are secreted and others of which are bound to cell surface heparin or

ECM structures. VEGF165 is an abundant isoform and also the most widely studied.

22 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation In healing wounds, VEGF is mainly produced by stromal fibroblasts and inflammatory leukocytes (Bao et al., 2009; Nissen et al., 1998). In addition to proliferation, VEGF also contributes to vascular maintenance by promoting survival and preventing apoptosis in endothelial cells. In addition, it increases vascular permeability, which is important in inflammatory reactions, as well as in the construction of new vasculature. The interplay between VE-cadherin and VEGF has been studied in an in vivo-model of VEGF-dependent angiogenesis: VEGF associated with VEGFR2 induced tyrosine phosphorylation on VE-cadherin, and the level of phosphorylated VE-cadherin was higher in angiogenic tissues, indicating activation of downstream signaling pathways leading to angiogenesis (Lambeng et al., 2005). The importance of VEGF signaling in development is demonstrated by lethality in a mouse model with a heterozygous knock-out of VEGF (Ferrara et al., 1996). VEGF exhibits a temporal pattern of expression in healing wounds (Bao et al., 2009; Brown et al., 1992; Kumar et al., 2009), and maximal levels of VEGF occur simultaneously with granulation tissue formation, approximately 3 to 7 days post- wounding (Brown et al., 1992). VEGF and VEGFR2 levels remain elevated long into the scar phase (Kumar et al., 2009), indicating the importance of VEGF signaling in later remodeling functions. Hepatocyte growth factor or scatter factor (HGF or SF) is another growth factor with an important role in the tissue repair process. Its main source is the stromal fibroblast, and it promotes various functions, such as cell motility, proliferation, morphogenesis, and matrix construction in several cell types, such as hepatocytes, endothelial cells, epithelial cells, melanocytes, and keratinocytes (Conway et al., 2006; Nakamura et al., 1986; Weidner et al., 1991). It is secreted in an inactive form, and activated by proteolytic cleavage in damaged tissues, mainly by the ser- ine protease HGF activator (HGFA) (Miyazawa et al., 1996). Known HGFA activators are the HGF-converting enzyme, as well as components of the coagulation cascade such as thrombin, urokinase-type plasminogen activator, tissue-type plasminogen activator, and blood coagulation factor XIIa (Conway et al., 2006; Mars et al., 1993; Miyazawa et al., 1996; Shimomura et al., 1993; Shimomura et al., 1995). HGF binds to its receptor MET (mesenchymal-epithelial transition factor), which is present mostly on epithelial cells, but also on endothelial cells (Birchmeier et al., 2003; Nakamura et al., 1995; Park et al., 1986; Sonnenberg et al., 1993). HGF is now also recognized as a promoter of angiogenesis and lymphangiogenesis and

Review of the literature | 23 functions by encouraging proliferation, migration, and sprouting of vascular and lymphatic endothelial cells (Bussolino et al., 1992; Kajiya et al., 2005). MET plays an important role in wound healing, since the presence of MET in keratinocytes is obligatory for successful wound re-epithelialization. MET signaling leads to activation of motogenic signaling pathways, actin polymerization, and the formation of focal adhesions and stress fibers, all required for cell migration (Chmielowiec et al., 2007). CXC-chemokines possessing the ELR motif (Glu-Leu-Arg motif), such as CXCL1-8, have angiogenic capabilities and stimulate endothelial cell chemotaxis (Strieter et al., 1995). All these chemokines bind to CXCR2, which is suspected to be the main pro-angiogenesis signal mediator (Keeley et al., 2008). Also ELR-negative CC-chemokines can also be angiogenic, for example CCL2 promotes angiogenesis in several in vivo models (Salcedo et al., 2000). Recent research has linked together COX-2 expression, chronic inflammation, and increased angiogenesis (Chang et al., 2004; Wang et al., 2005). The angiogenic effects of COX-2 can be grouped into three categories: induction of endothelial cell proliferation, enhancement of survival, and stimulation of cell adhesion and migration (Monnier et al., 2005). The COX-2 products, prostaglandins, have been shown to stimulate expression of VEGF and HGF (Liu et al., 1999; Matsunaga et al., 1994), and upregulation of COX-2 is known to be associated with HGF-promoted angiogenesis (Boccaccio et al., 2005). In addition to playing a major part in directing the general proinflammatory response, the transcription factor NF-κB has an important role in angiogenesis. It controls the transcription of many secreted signaling molecules, such as VEGF and several proangiogenic chemokines (Freter et al., 1996; Kiriakidis et al., 2003; Martin et al., 1997; Mauviel et al., 1992; Tabruyn and Griffioen, 2008), as well as the induction of COX-2 (Newton et al., 1997). In addition, it promotes the survival of endothelial cells (Scatena et al., 1998; Zen et al., 1999). On the other hand, recent results also indicate a possible inhibitory effect of NF-κB in tumor-related angiogenesis, as repression of NF-κB resulted in increased tumor vascularization (Kisseleva et al., 2006). Hence, it would seem that NF-κB can have two very different effects on angiogenesis.

24 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Role of stromal fibroblasts in angiogenesis

The stromal microenvironment plays a crucial role in supporting neovascularization in healing wounds. Several publications have demonstrated the importance of stromal fibroblasts as promoters of sprouting and lumen formation, both in in vitro and in vivo models (Hughes, 2008; Klagsbrun et al., 1976; Montesano et al., 1993; Nakatsu et al., 2003). Nevertheless, as fibroblasts are a heterogeneous group of cells, their capability for inducing endothelial sprouting and tubulogenesis seems to vary between strains (Klagsbrun et al., 1976; Montesano et al., 1993). It has become apparent that fibroblasts promote angiogenesis by secreting soluble factors such as prostaglandins (Wang et al., 1996; Zhang et al., 1998), cytokines (Brouty-Boyé et al., 2000; Lukacs et al., 1994), HGF (Weidner et al., 1991), and VEGF (Nissen et al., 1998). Fukumura et al. observed VEGF in granulation tissue around wounds, and they noted that the strong stromal VEGF signal originated mainly from fibroblasts (Fukumura et al., 1998). Fibroblast proximity is important for many endothelial cell functions. An in vitro model with reconstructed connective tissue showed that the presence of fibroblasts as well as ECM was necessary for tubulogenesis (Berthod et al., 2006), but in vivo the situation may be more complex. Nevertheless, fibroblasts are an important source of fibronectin and thereby could also promote angiogenesis by proxy (Hughes, 2008). The ECM protein fibronectin has been shown to promote angiogenic properties in endothelial cells including adhesion, growth, and survival, as well as the biological activity of HGF and VEGF (Rahman et al., 2005; Sottile, 2004; Wijelath et al., 2002). Hence, angiogenesis seems to be promoted by both ECM and its component fibronectin, as well as by matrix-producing fibroblasts.

Nemosis is a model for the stromal activation process of fibroblasts

In 2004, our laboratory discovered an experimental model for fibroblast activation. Depriving normal, primary, human foreskin fibroblasts of growth support by coating a U-well culture dish with a thin film of agarose resulted in their clustering into compact spheroidal structures. Clustering is not dependent on the coating material of the culture wells, as the nemosis response remained similar with poly(2-

Review of the literature | 25 hydroxyethylmethacrylate; polyHEMA) as coating material, and also occurred when we used the hanging-drop culture method (Bizik et al., 2004; Salmenperä et al., 2008). In our laboratory, 10-20 000 cells per spheroid are used, but early unpublished observations have indicated that this cell number is not critical for the nemosis response. Compact spheroids are formed within 24 hours (Figure 3), and around that time dramatic changes in mRNA and protein expression profiles take place. Simultaneously, the spheroids begin to decompose, displaying necrotic morphology. No upregulation of apoptosis markers such as Fas, Daxx, and activated caspase-3 is observed, nor is presence of fragmented DNA seen (Bizik et al., 2004). Initially this novel, programmed, necrosis-like cell death was thought to be a key element in nemosis, as a number of cells in the spheroid are dying. However, other cells seem to remain viable for at least five days, and as later research established similarities to stromal and cancer-associated fibroblasts, the actual role of such fibroblast death in vivo has not been determined.

3h 6h 12h

24h 48h 72h 96h Figure 3. Time-points depicting the clustering process of human primary dermal fibroblasts, resulting in a compact spheroidal structure.

Because of the apparent links to cell death as well as host defense and tissue repair, this fibroblast activation model was named “nemosis”, after Nemesis, the Greek goddess of retribution and inevitable consequences (Kankuri et al., 2005; see Table 2 for an overview and proposed timeline of nemosis).

26 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation 6 h 24 h 48 h 72 h 96 h Fibronectin Fibronectin HGF induction VEGF induction High levels production decrease of IL-1, -6, -11, GM-CSF, LIF COX-2 MMP-10 induction induction MT1-MMP, TIMP-1 induction Membrane leakage Necrotic morphology observed Prostaglandin induction Table 2. Overview and proposed timeline for nemosis. Induction of growth factors, MMPs and cytokines analyzed as secreted protein; fibronectin and COX-2 as expressed protein. GM-CSF: granulocyte macrophage-colony stimulating factor; LIF: leukemia inhibitory factor (Bizik et al., 2004; Kankuri et al., 2005; Kankuri et al., 2008; Räsänen et al., 2008; Salmenperä et al., 2008; Sirén et al., 2006).

Allowing fibroblasts to cluster into spheroids leads to the induction of COX-2 and the production of prostaglandins E2, I2, and F2α. COX-2 is induced throughout the spheroid and peaks around 60-72 hours. The selective COX-2 inhibitors indo- methacin and NS-398 were able to inhibit not only COX-2 activity and prostaglandin synthesis, but also the expression of COX-2. Spheroid formation was not affected. The expression of the non-inducible, ubiquitous COX-1 persists long into the process (Bizik et al., 2004). We have been unable to find any markers for oxidative stress or hypoxia in the spheroid. Also, we have noted that the induction of the hypoxia-inducible COX-2 is uniform and not restricted to the center of the spheroid, as could be the case in hypoxia. Interestingly, the nemosis response is not constant between fibroblast strains. Measured as COX-2 induction and the production of VEGF and HGF, along with cancer-associated fibroblast markers such as fibroblast-specific protein 1, α-SMA, and fibroblast activation protein, it varies slightly among primary human fibroblast strains, as well as between normal and cancer-associated fibroblasts, underscoring the heterogeneity of the fibroblast (Räsänen et al., 2009).

Review of the literature | 27 Early steps in nemosis of fibroblasts: interactions with the extracellular matrix

The ECM protein fibronectin mediates cluster formation during the compaction stage, as demonstrated by the fact that fibronectin knock-out or RNAi-silenced cells form loose clusters instead of compact spheroids (Salmenperä et al., 2008).

Fibroblasts are known to bind fibronectin mainly via α5β1 and αvβ1-integrins, and the importance of these integrins in the initial compaction of single, scattered cells into a firm spheroid is illustrated by the delayed spheroid formation that occurred

when α5 or β1-integrins were functionally blocked, resulting in a loose spheroid for the first 12 hours (ibid). A similar effect was seen if the fibronectin in the cells pos- sessed a mutated integrin binding site (RGE (Arg-Gly-Glu)-amino acid sequence

instead of RGD (Arg-Gly-Asp)) (ibid). Blockage of αv and β1, as well as specific blockage of the fibronectin RGD site, resulted in the loss of COX-2 induction, indicating a role for fibronectin in nemotic activation of fibroblasts (ibid). In conclusion, nemotic activation of the clustering fibroblasts seems to result from fibronectin- integrin interactions that take place during the initial compaction stage . Proteolysis is an important event in tissue remodeling during tissue repair (Gill and Parks, 2008). Nemosis also leads to an upregulated expression of extracellular matrix metalloproteinases MMP-1, -10, and membrane type-1-MMP (MT1-MMP), and the latter two are also secreted in higher amounts. Furthermore nemotic fibroblasts display proteolytic activity caused by MMPs (Siren et al., 2006). Many MMPs are known to modify inflammatory responses by cleaving and processing cytokines and growth factors (Van Lint and Libert, 2007). Moreover, several MMPs, such as MMP-10, are also implicated in wound healing (Gill and Parks, 2008; Rechardt et al., 2000), whereas the expression of the soluble form of MT1-MMP, known to be produced by fibroblasts, can be upregulated in inflammatory conditions (Li et al., 1998; Maisi et al., 2002). Plasminogen activation and an increased generation of plasmin also takes place during nemosis, and these processes are mainly regulated by exposure to α-enolase on the cell surface (Bizik et al., 2004). As MMP and plasmin inhibitors did not affect spheroid decomposition significantly, it was suggested that proteolysis is targeted to the surrounding cells and not to the nemotic fibroblasts themselves (Siren et al., 2006).

28 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Influence of nemotic fibroblasts on epithelial cells

During wound repair, one of the main goals of the stroma is to support healing processes such as dermal closure and remodeling (Braiman-Wiksman et al., 2007). Keratinocytes are crucial in early events of wound repair, and their migration and proliferation is needed to close the wound with a new protective layer. The HaCaT cell line is a spontaneously immortalized skin keratinocyte cell line (Boukamp et al., 1988), which has been shown to modify nemotic activation of fibroblasts, determined as abolished COX-2 expression. In a model with an increasingly malignant HaCaT cell panel (immortalized, benign, malignant, and metastatic lines) representing squamous cell carcinoma progression (Mueller and Fusenig, 2002), the more malignant types were shown to behave in an opposite manner by increasing the expression of COX-2 and production of growth factors (Räsänen et al., 2008). The more malignant HaCaT cell types induce the production of keratinocyte growth factor, HGF, and VEGF in nemotic fibroblasts (ibid). Nemotic fibroblasts in turn increase the proliferation and motility of HaCaT cells (Räsänen and Vaheri, 2010). These signaling processes indicate a paracrine cross-talk between epithelial cells and nemotic fibroblasts. Bone marrow mesenchymal stem cells have also been shown to undergo nemosis, during which they induce their COX-2 expression and produce high levels of HGF. Nemotic bone marrow mesenchymal stem cells induce keratinocyte scratch wound closure in vitro in an HGF/MET/phosphoinositide-3-kinase-dependent manner (Peura et al., 2009). Since bone marrow mesenchymal stem cells are present in skin wounds (Badiavas et al., 2003), inducing keratinocyte proliferation through nemosis suggests interesting possibilities for wound healing and tissue regeneration.

Influence of nemotic fibroblasts on malignant cells

Many inflammatory processes are also occurring during tumorigenesis, and the link between inflammation and cancer is being intensely investigated (Mantovani et al., 2008). Indeed, inflammation has been proposed to be the seventh hallmark of cancer (Colotta et al., 2009), in addition to the six generally acknowledged steps

Review of the literature | 29 in tumorigenesis as laid out by Hanahan and Weinberg (Hanahan and Weinberg, 2000). Tumors seem to be capable of inducing and redirecting tissue healing responses for the benefit of cancer progression, among other things by recruiting inflammatory cells and fibroblasts that together make up the stroma surrounding the tumor. These stromal cell types secrete large amounts of cytokines, chemokines, metalloproteases and growth factors that promote further growth and develoment of the tumor (Tlsty and Coussens, 2006). Some types of tumors consist mostly of stromal cells including fibroblasts, which have been shown to promote tumorigenesis. These fibroblasts are also called cancer-associated fibroblasts or CAFs (Schurch et al., 1981). Activated, nemotic fibroblasts produce and secrete abundant amounts of HGF, which promotes the invasive behavior of melanoma and carcinoma cells that express the functional HGF receptor MET (Kankuri et al., 2005). On the other hand, MET expression by nemotic fibroblasts was shown to decrease, which suggests that the HGF production is probably for paracrine signaling purposes. MET-negative leukemia cells respond differently: when stimulated by nemotic fibroblasts they undergo growth arrest and differentiation to a dendritic phenotype (Kankuri et al., 2008). The studies above show that nemotic fibroblasts are able to direct changes in phenotype in tumor cells, as well as changes in their behavior. Interestingly, conditioned medium from Bowes melanoma cells induced spontaneous clustering of fibroblasts in vitro, forming clumps much like artificially formed fibroblast spheroids (Kankuri et al., 2005). Such an effect suggests parallels to the cross-talk taking place between tumor cells and CAFs.

30 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Aims of the study

This study was undertaken to characterize fibroblast nemosis and the different ways it could promote and regulate proinflammatory and proangiogenic properties. In addition, this study attempted to shed light on how nemosis, or its possible in vivo counterpart, could contribute to the emergence and progression of inflammation, with possible implications for inflammation-related tumorigenesis, another aspect of nemosis investigated by our group. The study was divided into three sub-projects, which together constitute this doctoral thesis:

1. Do nemotic fibroblasts produce chemoattractants and call leukocytes?

An initial microarray analysis conducted to profile the nemotic response of -fi broblasts on the gene expression level showed that nemotic activation led to upregulation of many chemokines. We aimed to validate the chemokine profile and further investigate its effect on leukocytes. We hypothesized that nemotic fibroblasts secreted chemokines and that they would promote leukocyte migration. Nuclear factor-κB (NF-κB) is an important component regulating cytokine gene expression and other mechanisms leading to a proinflammatory response. There- fore, we hypothesized that NF-κB was activated in nemotic fibroblasts and regulates chemokine production.

2. Do nemotic fibroblasts activate endothelial cells?

Since nemotic fibroblasts were shown to display inflammation-promoting and leukocyte-attracting properties, we hypothesized that they were also capable of activating endothelial cells and leukocytes, resulting in enhanced leukocyte adhesion on endothelium. We also hypothesized that activated, nemotic fibroblasts could increase the permeability of the endothelial layer, aiding extravasation of leukocytes into tissues. We aimed to examine the influence of nemotic fibroblast medium on leukocyte adhesion onto endothelial monolayers, as well as screen for

Aims of the study | 31 surface markers on the endothelial cells. Furthermore, we aimed to investigate permeability-related changes in the adherens and tight junction proteins of endothelial cells in a monolayer.

3. Do nemotic fibroblasts promote angiogenesis?

Our group found that nemotic fibroblast spheroids produced and secreted large amounts of HGF, a growth factor which promotes cell proliferation and angiogenesis. Furthermore, our group has shown that HGF produced by nemotic fibroblast spheroids promotes scattering and invasiveness of c-Met-expressing human cancer cells (Kankuri et al., 2005). Production of VEGF was also elevated in nemotic spheroids when compared to fibroblast monolayer cultures (Räsänen et al., 2008). These clues led us to investigate the angiogenic potential of nemotic fibroblasts.

32 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Materials and methods

The materials and methods used in this study are summarized below. Detailed descriptions, including parallel experiment repetition numbers and statistical analyses, can be found in the original publication(s) indicated by the Roman numeral in question.

Cell isolation, culture, and initiation of nemosis in fibroblasts (I, II, III)

Human cell types used in this study are given in Table 3 (see following page). All cells were routinely cultured in a humidified, 5% CO2 atmosphere. Adherent cells were detached by trypsinization. Neutrophils were freshly isolated from whole blood on the day of the experiment and diluted with trisodium citrate and BSA, layered over isotonic Ficoll-Paque dextran (Amersham Biosciences), and centrifuged with density gradient separation (Roos and de Boer, 1986). Erythrocytes in the granulocyte fraction were removed by lysis with isotonic NH4Cl, and the remaining granulocyte fraction was washed. The purity of neutrophils was >97%. The nemotic activation process of fibroblasts was initiated by allowing 104 cells to cluster upon a nonadhesive surface (96-well U-bottom plate coated with 0.8% low melting-point agarose), thus forming a spheroid structure. Monolayer cultures of fibroblasts were used as a control, and these were directly seeded onto flat-bottom plates at an identical concentration. In chemotaxis experiments and the assay for CCL5 production (I) media were concentrated: spheroids were harvested at 24 h, concentrated to 48 spheroids/ml and grown in 12-well flat-bottom plates until further use. Monolayer cultures were directly seeded onto 12- or 96-well plates at a corresponding concentration.

Materials and methods | 33 Cell type name Description Publication(s) Source HFSF Primary foreskin I In-house fibroblast isolation CRL-2088 Primary foreskin II, III American Type fibroblast Culture Collection PMN Primary poly- I Freshly isolated, morphonuclear in-house neutrophil THP-1 Leukemic mono- I, II American Type cyte cell line Culture Collection HAL Primary adult lung I In-house fibroblast isolation HES Primary embryonal I In-house skin fibroblast isolation MCR-5 Primary embryonal I American Type lung fibroblast Culture Collection HUVEC Primary umbilical II, III In-house, by vein endothelial Dr. H. Koistinen, cell isolation University of Helsinki HMEC-dNeo Primary neonatal III by Dr I. Virtanen, dermal University of microvascular Helsinki, endothelial cell originally pur- chased from Clonetics/Lonza Table 3. Cell lines (all human) used in this study.

Microarray experiments and analysis (I)

Total RNA from corresponding fibroblast spheroid and monolayer samples was extracted by Trizol Reagent (Invitrogen) and purified using the RNeasy kit (QIA- GEN). The quality of the mRNA was assessed using a Bioanalyser 2100 (Agilent Technologies). Biotinylated cRNAs were prepared according to manufacturer’s instructions (Affymetrix), fragmented, and hybridized to Affymetric HG-U133A GeneChips using standard conditions in the Affymetrix fluidics station. Scanned arrays were analyzed by Affymetrix Microarray Suite 5.0 software by omitting obvious artifacts and outliers, scaling raw expression intensity, and adjust-

34 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation ing mean signal of probe sets to a user-specified 500. The gene expression profiles were compared using GeneSpring™ software and normalized both “per chip” and “per gene”, as well as filtered for “molecular function: chemokine”. The resulting list was reduced to those classified as chemokines by the IUIS/WHO Subcommittee on Chemokine Nomenclature (Thorpe et al., 2003).

Quantification of growth factors and chemokines in conditioned fibroblast medium (I, III)

Conditioned media were collected from nemotic fibroblasts and monolayers at different time points. Quantification of CCL2, CCL3, CCL5 (R&D Systems), CXCL8 (Sanquin Reagents), and VEGF (R&D Systems) in the conditioned medium was carried out by ELISAs according to the manufacturers’ instructions. The media for the CCL5 assay were concentrated as described on page 33.

Chemotaxis assay (I)

The chemotactic activity of conditioned fibroblast media was assayed for primary neutrophils and the monocytic THP-1 cells using a Boyden chamber setup. Leukocytes were labeled with fluorescent Calcein-AM (Molecular Probes/Invitro- gen) and placed in the upper chamber of a Transwell plate system (Costar/Corn- ing). Conditioned nemotic fibroblast or monolayer medium was added to the lower well, and leukocytes were allowed to migrate through a porous membrane (pore size 3 µm) in cell culture conditions. Transmigrated cells in the bottom well as well as cells attached to the membrane were lysed with an HTAB-based buffer and the fluorescence measured with a fluorometer (Wallac Victor2 1420, PerkinElmer).

Materials and methods | 35 Determination of IκB levels and NF-κB activity (I)

IκB was detected by immunoblotting: whole-cell samples were processed for SDS-PAGE and the separated protein bands transferred to a nitrocellulose membrane, which was then probed with anti-IκB antibody and a secondary antibody with an HRP conjugate. The ECL signal was detected on X-ray film by the Super- Signal West Pico substrate kit (Pierce Biotechnology). To detect changes in protein levels, densitometric analysis of the bands was carried out using a ChemiDoc XRS instrument and Quantity One software (Bio-Rad). Band intensities were normalized to actin at corresponding time points to yield relative values. NF-κB levels were analyzed from nuclear extracts. The extraction was performed using the Nuclear Extraction Kit by Cayman Chemical, and nuclear protein concentrations were assessed by the DC Protein Assay kit (Bio-Rad). NF-κB activity was analyzed by an enzyme-linked, immunosorbent NF-κB (p65) Transcription Factor Assay Kit (Cayman Chemical). Briefly, nuclear extract samples were normalized for protein load and the total volume incubated with NF-κB consensus sequence DNA pre-bound to a sample well plate, after which a p65-specific antibody and an HRP conjugate were added. Finally, the relative NF-κB activity was detected by a colorimetric readout.

Analysis of tight and adherens junctions on endothelial monolayers (II)

HUVECs were grown on gelatin-coated coverslips to confluence for 96 h, followed by incubation with conditioned spheroid or monolayer medium. Coverslips were collected at various time points, fixed with paraformaldehyde and permeabilized with Triton X-100 (VE-cadherin, ZO-1), or fixed with acetone (claudin-5), and stored in PBS at +4°C until further use. Samples were stained for immunofluorescence with primary antibodies (ZO-1, VE-cadherin, claudin-5) and conjugated secondary antibodies (Alexa-488). Nuclei were stained with Hoechst 33342 dye. Finally, the coverslips were mounted on glass slides with mounting medium (SlowFade Gold, Molecular Probes/Invitro- gen). Images were captured with a Zeiss Axioplan2 microscope (Carl Zeiss Micro-

36 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Imaging) connected to a Hamamatsu CCD camera (Hamamatsu Photonics).

Static leukocyte adhesion assay (II)

HUVECs were plated on 24-well plates and grown to confluence for 96 h followed by stimulation with either monolayer or spheroid medium for 4 hours. GFP- expressing THP-1 cells were added, and the samples were incubated for 2 hours in cell culture conditions with or without function-blocking anti-β2-integrin antibodies (clone E74; the kind gift of Dr. C. Gahmberg, University of Helsinki). Nonadher- ent cells were washed away, the remaining cells were lysed, and GFP fluorescence was measured directly from the wells with a fluorometer.

Scratch-wound repair assay and live-cell imaging (III)

For the scratch-wound repair assay, endothelial cells were seeded on multiwell plates and grown to confluence, after which they were scratch-wounded with a pipette tip. The growth medium was replaced with conditioned nemotic fibroblast or monolayer medium, with or without function-blocking anti-HGF or anti-VEGF antibodies (R&D Systems). Cells were fixed with paraformaldehyde, permeabilized with Triton X-100, and stained with Hoechst dye 33342 for nuclear DNA visualiza- tion. Samples were imaged with an Olympus CKX41 inverted microscope, a UPlan Fluorite phase-contrast objective, and an Olympus DP12 CCD camera system. Analysis of the images was carried out with custom-made software (Molecular Imaging Unit Core Facility at Biomedicum Helsinki, Faculty of Medicine, Univer- sity of Helsinki, Finland) on a Matlab platform (MathWorks Inc.), yielding a relative value representing the cell-free area for each image. The scratch-wound closure process was also followed in real-time with a Cell- IQ cell culturing platform (Chip-Man Technologies), equipped with a phase-contrast microscope (Nikon CFI Achromat) and camera. The system was computer- controlled by Imagen software (Chip-Man Technologies). The experimental set-up was identical to the scratch-wound assay, and images were captured at 30-minute intervals for 24 hours. Analysis was carried out with ImageJ software (Rasband,

Materials and methods | 37 1997), using the Manual Tracking plug-in. Images were stacked and the movements of eight cells on the wound edge per image were tracked.

Endothelial sprout formation assay (III)

Endothelial cells were plated on polymerized Growth Factor Reduced Matrigel® basement membrane matrix (BD Biosciences) and covered with conditioned nemotic fibroblast or monolayer medium with or without function-blocking anti-HGF and anti-VEGF antibodies. Plates were incubated for 24 hours, after which phase- contrast images were randomly captured with an Olympus CKX41 inverted microscope connected to a DP12 CCD camera (Olympus). Image analysis was carried out with ImageJ software. Sprout length was assessed as the distance between two branching points and scaled back to micrometers.

Proliferation assay (III)

HUVECs were plated onto 6-well plates (50 000 cells/well) and cultured for 24 h. The culture medium was then replaced with conditioned nemotic fibroblast or monolayer medium and incubated for a further 48 hours. Phase-contrast images were captured at various time points and analyzed by counting the number of cells per image. Counting was carried out with ImageJ software, using the Particle Analy- sis/Cell Counter plug-in.

38 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Results and discussion

The main results of the study are presented and discussed here. For a more detailed description and discussion the reader is referred to the original research publications indicated with corresponding Roman numerals.

Nemotic fibroblasts exhibit a proinflammatory response (I, II, IV)

Research has unveiled a new functional role for fibroblasts. Instead of only playing a structural role in the construction and maintenance of ECM, fibroblasts are now also seen as modulators of the inflammatory defense response in tissue (Ba- glole et al., 2006; Flavell et al., 2008). Activated fibroblasts in the inflammatory stroma are known to produce an array of chemokines that attract leukocytes, and to activate and uphold the inflammation response through an array of secretory molecules such as cytokines, matrix metalloproteinases, and growth factors.

Production of chemotactic cytokines and attraction of leukocytes

Nemosis is a model for stromal activation of fibroblasts. Upon clustering, extensive changes take place, and the fibroblasts convert to an increasingly secretory, proinflammatory phenotype, producing among other factors prostaglandins, proteolytic enzymes, and growth factors (Bizik et al., 2004; Kankuri et al., 2005; Räsänen et al., 2009; Sirén et al., 2006; reviewed in (IV)). The proinflammatory response is further enhanced by the production of chemotactic cytokines, as described in this study (I). Nemotic fibroblasts upregulate the mRNA levels of several chemokines (Table 1). They also secrete these chemokines in significantly higher amounts than fibroblasts in monolayer cultures already at 36 h after cell seed- ing. Microarray analyses provided a set of chemokines with over five-fold increased expression compared to monolayer cultures, with top fold-increase values of >100 (CCL3 and -5). The chemokines CXCL1-3 were not studied further because of their double nature as growth factors and chemoattractants, and CCL4 because of its low signal intensity level. The remaining chemokines, CCL-3, -5, and CXCL8,

Results and discussion | 39 were also secreted at significantly higher levels by the nemotic fibroblasts than by monolayer cultures after 72 hours, with XCXL8 levels reaching as high as 40 ng/ml in the standard set-up of 104 fibroblasts seeded in 150 µl medium. Only CCL2 did not follow the pattern; microarray experiments showed a higher concentration of mRNA in nemotic fibroblasts, but secreted protein levels remained lower than in monolayers. In endothelial cells it seems that CCL2 could be stored intracellularly but not secreted (Øynebråten et al., 2004; Øynebråten et al., 2005), and a similar phenomenon could perhaps also take place in nemotic fibroblasts, although this was not further investigated. The production of cytokines seems to be an intrinsic property of fibroblasts both in vitro and in vivo, in that a basal level of cytokine production is constantly maintained. Fibroblasts in inflamed areas are known to display an altered chemokine profile (Brouty-Boyé et al., 2000; Hogaboam et al., 1998; Lukacs et al., 1994). Pro- duction of chemotactic cytokines indicates a paracrine effect, and the chemokine profile produced by nemotic fibroblasts points to leukocytes as target cells. Indeed, in this study we showed that conditioned medium of nemotic fibroblasts was able to attract both neutrophils and monocytic THP-1 cells, with an increase of 1.4 and 6.4-fold, respectively (I). Chemotaxis was partially dependent on CXCL8 (maximal inhibition 24% with antibodies) with regard to neutrophils, and CCL3 and CCL5 with regard to THP-1 cells (maximal inhibition of the CCL3 and CCL5 receptor CCR1 was 38%). CCL5 functions as an activator and chemoattractant of T cells (Schall et al., 1990). Furthermore, nemotic fibroblasts have been shown to produce the cytokines IL-1, -6, -8, -11, leukemia inhibitor factor, and granulocyte-macrophage colony-stimulating factor, as well as to induce dendritic differentiation of leukemic cells such as THP-1 (Kankuri et al., 2008), which widens the profile for the effects of fibroblast spheroid-mediated inflammatory activation of leukocytes. Expression of many cytokines is regulated by the transcription factor NF-κB, which was also found to be activated in nemosis, as shown in this study (I). The two components of the classical NF-κB pathway, p65 and p50, were present in nemotic fibroblasts with stable expression levels throughout the nemosis progression from 0 to 96 hours. The endogenous inhibitor IκB was degraded in a fluctuating pattern with dips at 36 and 48 hours, suggesting periodic boosts in NF-κB activity. Func- tionally, NF-κB was more active in nemotic fibroblasts than in monolayer cultures at the aforementioned time points. Fibronectin is produced especially during the

40 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation first 24 hours of nemosis (Salmenperä et al., 2008) and is known to activate NF- κB, at least in epithelial cells (Kolachala et al., 2007). If a similar activation exists in nemotic fibroblasts, a possible feed-back loop may take place during the compaction stage. It has been proposed that chronic inflammation happens when immunoactive fibroblasts fail to switch off, and that fibroblasts would thus play an important role in the attenuation of inflammation (Buckley et al., 2001). The mechanism for attenuation in fibroblasts is unknown, but according to some studies, fibroblast RelB is capable of stabilizing IκB, which in turn switches off NF-κB and thus also the plethora of proinflammatory molecules produced by the fibroblasts (Lo et al., 1999; Xia et al., 1997). Constitutively active NF-κB is often reported in inflammatory pathologies (Tak and Firestein, 2001); hence, active NF-κB in stromal fibroblasts could point to it having a possible role in persisting inflammation. Leukocytes play an important role in another mechanism for the attenuation of inflammation. The production of prosolving lipid metabolites downregulates cytokine expression, regulates macrophage phagocytosis, and inhibits neutrophil infiltration (Serhan et al., 2008). These lipid mediators are mainly produced by mucosal and vascular cells, as well as leukocytes and platelets. COX-2 is required for the production of many of these mediators, by way of acting as a switch in en- zymatic function that results in production of anti-inflammatory molecules such as lipoxins, protectins, and resolvins (ibid). PGE2, a COX-2 end product traditionally viewed as proinflammatory, has been shown to induce lipoxin production in leukocytes (Levy et al., 2001), suggesting a feed-back mechanism that promotes homeostasis. Fibroblasts in nemosis seem to share properties with inflammation- promoting fibroblasts, but as two of the characteristics of nemosis are COX-2 upregulation and PGE2 production, the possible implication of this for the resolution process of inflammation should be considered. All experiments on nemotic fibroblasts in this study have been conducted in comparison with fibroblast monolayer cultures, a traditional way of observing mesenchymal cells. The suitability of monolayer cultures in comparison to cultures with spheroidal structures can be discussed, especially in the light of how cells react to 2D versus 3D environments. Cukierman et al. have performed extensive studies on fibroblasts in 3D cultures, showing clear differences in fibroblast phenotype when grown in different lattices such as collagen and ECM (Cukierman et al.,

Results and discussion | 41 2001; Cukierman et al., 2002). Fibroblasts in stressed collagen and ECM are known to proliferate, whereas a relaxed collagen matrix causes quiescence. In contrast, at least some of the nemotic fibroblasts seem to suffer a necrosis-like fate. Adhesion

of 3D-grown fibroblasts was shown to be dependent on the integrin α5β1, and cel-

lular responses such as migration and proliferation on α5 (Cukierman et al., 2001).

Both α5 and β1 regulate clustering in the nemosis model (Salmenperä et al., 2008). While 3D cultures and nemosis may share properties, proliferation arrest, along with the induction of growth factors, prostaglandins, cytokines, and MMPs seen in nemosis, suggest that nemosis is not identical to fibroblasts grown in 3D lattices. Still, a careful study of spheroid structures grown in different lattices may yield information indicating whether one of them provides a better control environment than conventional monolayer culture.

Activation of endothelium and modulation of its permeability

Leukocytes, especially monocyte-macrophages, are a crucial part of the first line of defense against harmful agents. They also play an important part in directing inflammatory and tissue repair responses. For these circulating blood cells to home to inflamed tissue they must be alerted by signals from the site of inflammation that reach into the vascular compartment (Chavakis et al., 2009; Granger and Kubes, 1994; Muller 2009). The task of alerting leukocytes is carried out by the vascular endothelium. The endothelium lines vessel walls and provides a body- wide sensory system for damage alerts from within tissue. Only activated endothelium is known to be able to bind leukocytes, and such activation can be induced by stromal components such as fibroblasts (Maiellaro and Taylor, 2007; Zimmerman et al., 1999). This study shows that activated, nemotic fibroblasts promote adhesion of monocyte-like THP-1 cells onto a HUVEC monolayer by 82%. The adhesion is

partly β2-integrin-dependent, as a blocking antibody inhibited adhesion by 15%.

A blocking antibody against β1-integrin had no inhibitory effect (II). Firm adhe-

sion of leukocytes is part of the extravasation process, and mediated through β2- integrins on the leukocyte and ICAM-1 on the endothelium (Gahmberg et al., 1997, Gahmberg et al., 2009; Lawson and Wolf, 2009). However, the results suggest that

42 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation β2-integrin-dependent adhesion is not the major adhesion mechanism in nemosis- mediated leukocyte adhesion to endothelium. Circulatory leukocytes must cross the endothelial barrier of the vascular wall in order to access the underlying injury site. Access is granted by activated endothelial cells that downregulate the expression of cell-cell junctions, thus locally increasing the permeability of the vessel structure. This study describes an effect of nemotic fibroblasts on endothelial cell junctionsin vitro (II). Conditioned medium from nemotic fibroblasts causes removal of the tight junction component claudin-5 and adherens junction component VE-cadherin, both membrane-bound proteins, from endothelial monolayer cell-cell junctions after 6 hours of stimulation with nemotic fibroblast medium. Claudin-5 and VE-cadherin interact with ZO-1 on the cytosolic side, which was also downregulated after 24 hours. While claudin-5 and VE-cadherin levels seemed to diminish, ZO-1 expression seemed to shift to the cytoplasm, indicating internalization and higher persisting levels of protein within the cells. Interestingly, claudin-5 was inconsistently expressed in some cells but not others, suggesting heterogeneity in the HUVEC cells. β-cadherin, another cytosolic linker protein, was not affected. These results indicate down-regulation of adherens- and tight junctions, resulting in a loosening of the HUVEC monolayer, a sign of proinflammatory activation. Increased endothelial permeability is a mechanism required for both leukocyte extravasation and angiogenesis, and both phenomena are required in wound repair. VEGF is also produced in nemosis, as shown in this study (III), in line with another publication from our group (Räsänen et al., 2008). Even if VEGF has been characterized as a master regulator of angiogenesis, it was first identified as a vascular permeability factor (Senger et al., 1983). Intracellular signaling through VEGF association with VEGFR2 has been shown to induce tyrosine phosphorylation on VE-cadherin in an in vivo-model of VEGF-dependent angiogenesis (Lambeng et al., 2005), suggesting that the VEGF stimulus causes reduced permeability of the endothelium. The study by Lambeng et al., along with another study by Gavard and Gutkind (Gavard and Gutkind, 2006) proposes that VEGF-stimulation activates VE- cadherin internalization by way of Src, resulting in VE-cadherin phosphorylation, subsequent recruitment of β-arrestin2, and receptor internalization. Microvascular integrity is also maintained by the promotion of new integrin connections through the connection of VEGFR2 and Src by VE-cadherin. This mechanism plays a role in

Results and discussion | 43 fluid shear stress mechanosensing of endothelial cells (Tzima et al., 2005). Because fibroblasts in nemosis produce VEGF and modulate endothelial VE-cadherin (II, III), the aforementioned cooperative interactions of VEGF and VE-cadherin may also regulate the endothelial permeability-modulating response of nemosis. Fibroblasts

in nemosis also produce PGE2, a molecule known for its leakage-inducing proper-

ties (Crunkhorn and Willis, 1971). COX-2 and its product PGE2 are upregulated by fibronectin (Han et al., 2004), which in turn has a crucial influence on the onset of nemotic activation of fibroblasts (Salmenperä et al., 2008). A recent hypothesis suggests that vascular inflammation can also take place as a result of outside-in signaling originating from the microenvironment, where activated, proinflammatory fibroblasts and leukocytes agitate the endothelium and thus indirectly also circulatory leukocytes (Maiellaro and Taylor, 2007). Such regu- latory behavior of the surrounding tissue may have implications in cardiovascular disease. In this study, the experiments on endothelial activation have been conducted using human umbilical vein endothelial cells (II, III). It can be argued that differences exist between arterial and venous endothelial cells (Sato, 2003; Wang et al., 1998) and that the physiological location for most inflammatory endothelial activation and leukocyte transmigration is the post-capillary venule (Granger and Kubes, 1994). Nonetheless, HUVECs have been shown to work remarkably well as an in vitro model for such functions, as most surface markers have been initially discovered in HUVECs and later confirmed in vivo, at a physiologically relevant location (Zimmerman et al., 1999). Yet, the extent of HUVEC resemblance to microvascular cells and the tissue-specific heterogeneity of endothelial cells can be debated. Certainly, differences do exist, for example in the expression of chemokine receptors (Gerber et al., 2009), but because many inflammatory activation-specific markers are present in both microvascular and macrovenous endothelial cells, it would seem probable that HUVECs perform decently as an easily handled, general in vitro endothelial cell model.

44 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Fibroblast nemosis has angiogenesis-promoting properties (III)

Wound repair requires a complex scaffold or microenvironment, consisting of several cell types, ECM structures, and signaling processes that aid in the recon- struction of the tissue. Angiogenesis is an important part of the formation of such a scaffold, and the vascular-forming response of endothelial cells is regulated by the tissue microenvironment (Langenkamp and Molema, 2009).

Nemotic fibroblasts promote angiogenic properties of endothelial cells

Previous work by our group revealed that nemotic fibroblasts are a source of proangiogenic growth factors such as HGF and VEGF (Kankuri et al., 2005; Räsänen et al., 2008), leading us to further investigate the potential proangiogenic role of nemosis. In this study we show that nemotic fibroblasts produce elevated levels of VEGF. Even if the increase in secreted VEGF is a moderate 2-fold, it contributes to the promotion of endothelial cell (HUVEC) sprouting and networking on basement membrane matrix, as shown by the effect of inhibitory antibodies against VEGF in conditioned nemotic fibroblast medium, as presented below. Fibroblasts undergoing nemosis are known to secrete abundant amounts of HGF of which at least 50% is in its active form (Kankuri et al., 2005). In this study we investigated the response of endothelial cells plated on basement membrane matrix after treating them with nemotic fibroblast or monolayer medium. The results show that HGF secreted into the medium also plays a role in the promotion of endothelial cell sprouting, as well as networking of the sprouts, together with VEGF. Inhibition of both HGF and VEGF by antibodies led to a maximal decrease of 39% in sprout number and 51% in the number of branching points, whereas sprout length remained unaffected. However, a combination of both function-in- hibiting antibodies did not result in a significantly enhanced inhibition of endothelial sprouting. This unexpected result could potentially be due to a convergence of downstream signaling routes, resulting in a maximum inhibition of sprouting when either HGF or VEGF is separately blocked. In this study we also investigated the effect of nemotic fibroblasts on human endothelial cells in a classical scratch-wound repair assay. Conditioned medium

Results and discussion | 45 induced significantly faster closure of both microvascular (HMEC-dNeo) and umbilical venous (HUVEC) endothelial monolayer scratch-wounds at 24 hours than medium from fibroblast monolayer cultures. The process was partly dependent on HGF, which inhibited closure up to 56%. VEGF, on the other hand, was not involved. Furthermore, we investigated whether this response was due to increased proliferation or migration of the endothelial cells. No signs of an induced prolif- eratory response could be detected, but we observed a moderate but statistically significant increase in the migration of single endothelial cells at the edges of the scratch-wound. Since HGF, like many other growth factors, acts as an endothelial cell motogen (Bussolino et al., 1992), its effect may be sufficient to close the wound. Interestingly, even if VEGF is known to induce endothelial cell proliferation, VEGF-containing nemotic fibroblast medium does not seem to promote proliferation of endothelial cells in vitro. Fibroblasts undergoing nemosis secrete an assortment of angiogenesis-inducing signaling molecules. Blocking one or two such molecules is evidently not sufficient to abolish angiogenic responses in endothelial cells. Yet, since a major source of HGF and VEGF in healing wounds is the activated fibroblast, the significant inhibitory -ef fects establish the importance of these factors in the angiogenesis-promoting role of nemosis. The HGF receptor MET, mostly expressed by epithelial and endothelial cells, can be activated by IL-1 and IL-6 (Moghul et al., 1994). These cytokines are also produced by fibroblasts in nemosis (Kankuri et al., 2008), suggesting a possible proactive manipulation of the HGF signal received by endothelial cells. Neutrophils are also known to be activated upon HGF stimulation, and they also are capable of storing abundant amounts of HGF (Jiang et al., 1992; Mine et al., 1998). It has therefore been suggested that ECM-bound and neutrophil-derived HGF provides the early source of HGF in tissue, whereas fibroblasts, once activated, are responsible for the sustained release of HGF. Interestingly, VEGF has been shown to induce CXCL12 (stromal cell-derived growth factor-1; SDF-1) in perivascular myofibroblasts, which in turn homes bone marrow-derived mononuclear myeloid cells to perivascular tissue, where they show a proangiogenic phenotype (Grunewald et al., 2006). Since stromal fibroblasts and inflammatory cells are both important sources of VEGF, it is appealing to think of activated fibroblasts as signal amplifiers, by both secreting VEGF and CXCL12, possibly by a feed-back loop. This in turn leads to the accumulation of monocytes and to further inflammatory progression.

46 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Other angiogenesis-promoting factors involved in nemosis

The expression of the proangiogenic chemokine CXCL8 has been found to be elevated in acute wounds (Engelhardt et al., 1998). The results presented here show that high levels of expressed and secreted CXCL8 is produced by nemotic fibroblasts, which suggests that the proangiogenic response of fibroblast nemosis may contain additional factors. Besides direct induction of angiogenesis, VEGF and HGF have been shown to further advance angiogenesis by activating signaling pathways in several cell types that contribute to the expression of CXCL8 (Dong et al., 2001; Nor et al., 2001), which could possibly also implicate a self-amplifying loop in nemotic fibroblasts. This study, along with other observations, concludes that COX-2 is robustly induced in the nemosis response of many types of fibroblast strains, together with the secretion of prostaglandins (Bizik et al., 2004; Räsänen et al., 2009). Expression of COX-2 has been linked to chronic inflammation and angiogenesis (Chang et al., 2004; Wang et al., 2005) and its upregulation is known to be associated with HGF- promoted angiogenesis (Boccaccio et al., 2005). This could have implications for nemosis-driven angiogenic responses. COX-2 and VEGF are known to be induced by hypoxia (Chida and Voelkel, 1996; Tuder et al., 1995), but we have been unable to find conditions of oxidative stress in nemotic fibroblasts. COX-2 is induced throughout the spheroid, and we have been unable to detect a necrotic or hypoxic center (Bizik et al., 2004). Therefore, we assume that the induction of VEGF, as well as other growth factors and cytokines produced, is also an intrinsic property of the nemosis process. Recently, the importance of NF-κB in angiogenesis has been recognized. This transcription factor functions as a hub, controlling the transcription of a plethora of cellular response elements, including several crucial for angiogenesis, such as VEGF (Kiriakidis et al., 2003; Tabruyn and Griffioen, 2008). Some angiogenic factors such as HGF are capable of activating NF-κB (Fan et al., 2005). Nemotic fibroblasts har- bor activated NF-κB, and they also and secrete abundant amounts of HGF, which could imply NF-κB-driven responses and feed-back loops, as well as have implications for the angiogenesis-promoting effect of nemosis. The production of fibronectin has been shown to be of importance in the early stages of nemosis, where it initiates cluster formation (Salmenperä et al., 2008).

Results and discussion | 47 Fibronectin also promotes endothelial cell adhesion, growth, and survival, as well as ECM-deposited HGF and VEGF activity (Rahman et al., 2005; Sottile, 2004; Wi- jelath et al., 2002). For new vasculature to form, space must be made by breaking down the ECM by proteolytic means. Fibroblasts are a major source of cellular fibronectin, and nemotic fibroblasts additionally secrete MMPs (Siren et al., 2006), providing a potential mechanism by which fibroblasts could aid angiogenesis indirectly (Hughes, 2008). This study demonstrates how nemotic fibroblasts influence the localization of the tight junction component claudin-5, the adherens junction component VE-cadherin, and the tight- and adherens junction adapter component ZO-1 in endothelial cell monolayers, causing them to be down-regulated at intercellular junctions. Vascular permeability is linked to neovascularization, and VE-cadherin is known to be of importance in angiogenesis by repressing endothelial proliferation, maintaining survival, and in such a manner stabilizing the vessel structure (Carmeliet et al., 1999; Caveda et al., 1996). Loosening of endothelial cell-cell contacts is required for sprouting and migration, the initial events performed by endothelial cells com- mitted to angiogenesis, and newly formed vessels are leaky until properly stabilized (Adams and Alitalo, 2007).

The dark side: inflammation-related damage and tumorigenesis

Inflammation has evolved to act as a driving force for wound repair in postnatal tissues. However, the inflammatory process can also be harmful. When a wound repair process fails to switch off the inflammatory phase, acute inflammation turns into chronic inflammation. Functionally altered fibroblasts are found in the stroma of chronically inflamed wounds, such as venous ulcers (Harding et al., 2005). Fi- broblasts are also abundant in scarring tissue (Desmouliere et al., 2005), as well as in interstitial fibrosis of the lung (Mornex et al., 1994) and kidney (González- Cuadrado et al., 1996), conditions which impair tissue and organ functionality. In fact, hypertrophic scarring seems to be an overreaction of stromal myofibroblasts (Desmouliere et al., 2005). Fibroblasts are also commonly present in granulomas, which have been described as cytokine factories (Mornex et al., 1994; Powell et al., 1999). A common factor in all these pathological conditions is the presence of

48 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation chronic inflammation, which drives the activation of fibroblasts. Fibroblasts attract leukocytes, and together these two cell types make up the main components of inflammatory stroma. Especially macrophages are known to be recruited by fibroblasts, and they further boost fibroblast immunoactivation, proliferation, and ECM production by additional chemokine production (Mrowietz et al., 1999). Such feed- back and cross-talk between cells in the stroma establishes a vicious cycle that augments the inflammatory response. Interestingly, inflammation is not a prerequisite for successful wound repair. Some early fetal tissues such as skin heal in a regenerative manner, since the acute immune response, and subsequently inflammation, are almost completely absent (Yannas et al., 2007). Such healing is scarless and includes perfect regeneration of the tissue surrounding the wound, including glands and hair follicles. It would seem that immune cells such as neutrophils and macrophages are not required for successful wound repair in a situation where infection is not a danger (Wilgus 2008). Fibroblasts are present in fetal wounds, but the matrix composition is different (Yannas et al., 2007) and the number of myofibroblasts is low (Estes et al., 1994). Bearing in mind that scar-forming repair has de facto evolved as the repair process of choice in post-natal mammals, consideration of the balance of positive and negative aspects in the wound healing process provides an interesting perspective on evolutionary advantages. In recent years it has become evident that chronic inflammation and inflammation-related responses are implicated in the initiation and progress of many types of cancer (Mantovani et al., 2008; Schafer and Werner, 2008). Chronic inflammation has also been shown to increase the risk of cancer at the site of inflammation (O’Byrne and Dalgleish, 2001). Classical examples are H. pylori-induced gastritis and gastric cancer, hepatitis B and C infection and liver cancer (Castello et al., 2010; Chemin and Zoulim, 2009), as well as inflammatory bowel disease and colorectal cancer (Balkwill et al., 2005). Solid tumors are surrounded by stroma, a scaffolding structure with resemblance to the inflammatory microenvironment found at sites of acute inflammation. A growth-supporting stroma seems to be an essential requirement for an advancing tumor (Mueller and Fusenig, 2002). Some types of cancer, such as scirrhous-type gastric carcinoma, are associated with extensive fibrosis and an abundance of fibroblasts (Semba et al., 2009). Fibroblasts found in the tumor stroma are in many

Results and discussion | 49 ways similar to inflammation-associated fibroblasts: they produce cytokines and growth factors required for tumor progression, whereas fibroblasts from normal, healthy tissue have been shown to suppress tumor progression (Olumi et al., 1999; Orimo and Weinberg, 2006; Pierce et al., 1982). These so-called cancer-associated fibroblasts (CAFs) have been shown to remain constantly activated, and often display markers of the myofibroblastic phenotype (De Wever et al., 2008; Hinz, 2007b; Kalluri and Zeisberg, 2006; Östman and Augsten, 2009). CAFs seem to be able to drive tumorigenesis starting from an early, hyperplasic stage, and continu- ing all the way through different stages of malignancy (Erez et al., 2010). Because of their important role in supporting the initiation and progression of cancer, CAFs are considered a possible therapeutic target (Kalluri and Zeisberg, 2006; Micke and Östman, 2004). Several factors produced by stromal fibroblasts also play roles in tumor progression. Activated fibroblasts attract leukocytes and support their maintenance, which has been suggested to indirectly aid tumorigenesis (Buckley et al., 2001). A certain subtype of macrophages has in particular been implicated in tumorigenesis (Silzle et al., 2003; Solinas et al., 2009). Chemokines, such as CXCL8 and CXCL5, are capable of promoting tumor cell proliferation (Põld et al., 2004), whereas CCL5 is implicated in tumor progression and metastasis-formation (Mrowietz et al., 1999). Cross-talk between malignant cells and fibroblasts seems to further stimulate tumor progression, illustrated by a study where melanoma cell lines stimulated the secretion of, among others, IL-6, CXCL8, and MCP-1 in fibroblasts, which then re- sponded by increasing the invasive potential of the melanoma cells (Li et al., 2009). Glu-Leu-Arg-positive CXCL chemokines are also known to promote angiogenesis, with implications for tumor vascularization (Gerber et al., 2009; Keeley et al., 2008; Strieter et al., 1995). The cytokine regulator NF-κB seems to be of crucial importance in many kinds of cancer, as it promotes the production of pro-tumorigenic cytokines and prevents apoptosis. These attributes make NF-κB a link between inflammation and cancer by producing proinflammatory agents, as well as by protecting cells with malignant potential from death (Greten et al., 2004; Karin, 2006). As rapidly growing structures, developing tumors require the infusion of nutrition and depend on neovascularization for development beyond a critical volume (Folkman et al., 1971). For historical reasons, many important discoveries in the

50 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation study of neovascularization have been made in a tumor environment. Tumor vascularization also provides a way for malignant cells to escape into the circulatory system, thus allowing metastasis. In contrast to fibroblasts isolated from normal tissue, CAFs originating from invasive breast carcinoma promote tumor vascularization (Orimo et al., 2005), and they have also found to be associated with microvascular proliferation (Zeine et al., 2009). CAFs vascularize an in vivo basement membrane plug more efficiently than fibroblasts from normal tissue, even without the presence of malignant cells (Erez et al., 2010). Effective vascularization of the tumor often correlates with poor prognosis. HGF has been linked to angiogenesis (Bussolino et al., 1992). High levels of the HGF receptor MET have been reported in many carcinomas, and such abundant expression seems to be essential for growth and survival of tumors (Shinomiya et al., 2004). Both HGF and MET are implicated in many types of cancer, and over- expression of either one results in tumorigenesis and aggressive metastasis (Birch- meier et al., 2003; Rong et al., 1994). VEGF, an important proangiogenic factor, was originally discovered in the context of cancer (Folkman et al., 1971). In tumors, the main source of VEGF is stromal inflammatory cells and fibroblasts (Fukumura et al., 1998). HGF and VEGF are known to cooperate by upregulating proinflammatory and angiogenic factors, as well as by promoting angiogenic responses of endothelial cells in a synergistic manner (Gerritsen et al., 2003; Wojta et al., 1999). Malignant keratinocytes seem to boost the HGF and VEGF produced by nemotic fibroblasts (Räsänen et al., 2008), suggesting enhanced growth factor production in a tumor environment. Since nemotic fibroblasts promote angiogenic responses in endothelial cells, the proangiogenic effect of nemotic fibroblasts may be enhanced by tumor cells. In conclusion, fibroblasts in the tumor stroma are an important part of the microenvironment that supports growth and progression of malignant tissue. Fi- broblasts undergoing nemosis display a variety of traits common to fibroblasts found in tumor stroma, such as the secretion of proteolytic enzymes, prostaglandins, growth factors, and cytokines. Recent research by Erez et al. has identified a proinflammatory gene signature, consisting of, among other, CXCL1, CXCL2, IL-6, COX-2, and NF-κB, in CAFs from various stages of dermal HPV-induced mouse skin carcinoma and human squamous cell carcinoma (Erez et al., 2010). All these markers are upregulated in nemosis. Similarities between inflammation- and cancer-

Results and discussion | 51 associated fibroblasts suggest that nemosis could be a model for not only inflammation- but cancer-associated stromal activation, as well.

52 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Concluding remarks

Activation of fibroblasts in the stroma of inflammatory tissue is an essential mechanism for promoting repair of damaged tissue. Related activation mechanisms have been found in the stroma of tumors. This study, along with other work by our group, has characterized clustering-induced fibroblast activation, or nemosis, and allowed us to link it with phenomena such as inflammation and tumorigenesis. We have illustrated a variety of responses and traits in nemosis that resemble those of stromal fibroblasts in inflammation, including cancer-associated inflammatory conditions (Figure 4).

Epithelial cells Inhibition of Tumor cells Promotion of nemosis proliferation (HGF) Promotion of nemosis Promotion of tumorigenesis (HGF, VEGF, MMP-10, MMP-14, PGs, plasminogen activation)

Promotion of inflammation Promotion of inflammation (MIP-1α, RANTES, IL-1β, IL-6, and angiogenesis IL-8, PGs) Nemotic fibroblast (HGF, VEGF, IL-8, PGs) cluster

Inﬂammatory cells Endothelial cells

Figure 4. Schematic model of signals produced by nemotic fibroblasts and possible paracrine cross-talk within the inflammation- and tumor microenvironment. Adapted from publication IV (Vaheri et al., 2009).

Currently, nemosis is an in vitro model that provides a way to investigate stromal activation in a controlled laboratory environment. The flexibility, versa-

Concluding remarks | 53 tility, and in-depth observation provided by in vitro studies is essential in understanding the mechanisms and potential impacts of nemosis. This study contributes to the foundation of research from which physiological and pathological implications can be derived, predicted, and understood. Clues for possible in vivo relevance scetch a picture of the inflammatory site, where an abundance of fibroblasts is often found. We may find the physiological counterpart of nemosis in interactions between these fibroblasts to be a mechanism for immunoactive awakening, in line with the sentinel cell theory of fibroblasts as an extension of the immunosurveil- lance system. In vitro, necrosis-like cell death occurs in the spheroid. However, in vivo the situation may be different and cell death may be delayed or lacking, due to intercellular signaling routes at the site of inflammation, prolonging the proinflammatory response of the fibroblast. In our model, proximity between fibroblasts is key to the activation process. Whether direct contacts without intervening ECM molecules are required for nemotic activation is as yet unclear. We know that fibronectin, an ECM protein produced by fibroblasts in abundance, plays an important role as a mediator for nemotic activation (Salmenperä et al., 2008), and that fibroblasts, specifically myofibroblasts, are known to make direct contacts in healing wound environments (Gabbiani et al., 1978). Connexin-43 is expressed in nemotic fibroblasts, but we have been unable to detect functional gap junctions between cells (our unpublished observations). Further work is required to elucidate how the nemosis model translates to a physiological situation, and whether it has an actual counterpart in tissue or if it is rather a convenient model to study stromal activation. Nemotic fibroblasts and myofibroblasts also share characteristics, but further profiling is required to determine a possible relationship between the two. Currently, in vivo detection of nemosis is complicated by the heterogeneity of fibroblasts and the lack of a precise tissue marker, as well as limitations in distinguishing actual cluster- or proximity-induced activation of fibroblasts in tissue. Most of our work has been conducted on primary human foreskin fibroblasts, in line with its potential relevance to skin wounds. In this study we have also examined other strains of human fibroblasts of skin and lung origin, both adult and embryonic, with the conclusion that these also show nemotic morphology, upregulation of COX-2, and secretion of CXCL8 (publication I; supplemental data). Notably, another study by our group compared paired strains of normal and can-

54 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation cer-associated fibroblasts and concluded that some variation in the nemosis response of fibroblasts does exist (Räsänen et al., 2009). This observation, along with unpublished observations on epithelial and endothelial cells that do not undergo nemosis, points to a conclusion that the general characteristics of nemotic behavior seem to be an intrinsic property of fibroblasts. Anchorage-dependent cells, such as epithelial and endothelial cells, are known to react to loss of growth support by losing their sense of polarity, resulting in an apoptotic form of cell death termed “anoikis”. Caspase activation and induction of apoptosis-related genes such as Fas and Baxx are key regulators of the anoikis pathway (Chiarugi and Giannoni, 2008). However, previous studies on nemosis have seen no upregulation of caspases or apoptosis-related proteins (Bizik et al., 2004). In the present study we also compared spheroid formation of HUVECs and HMEC- dNeo endothelial cells to fibroblasts, noting a rapid morphological decomposition and no induction of COX-2 (publication III, supplemental data), as expected of a cell type known to undergo anoikis in suspension conditions. We also seeded fibroblasts on flat-bottom, agarose-coated plates at low numbers and observed a spontaneous clustering of the cells, making a direct comparison of anoikis and nemosis of fibroblasts impossible (our unpublished results). Based on the results above we conclude that nemosis of fibroblasts is a pathway different from anoikis. The production of cytokines is strongly augmented in fibroblasts undergoing nemosis. So far we have not detected anti-inflammatory cytokines, and the chemokines produced include ones with both proinflammatory and proangiogenic effects. In addition, nemotic fibroblasts are also a rich source of growth factors, some with angiogenic properties, especially HGF and VEGF. Inducing nemosis in primary fibroblasts can be considered a means of directing inflammation and inducing angiogenesis, thus promoting a wound healing response and subsequent tissue regeneration in a controlled fashion. In conclusion, nemosis can be seen as an in vitro model that facilitates careful examination of the activation process of stromal fibroblasts, or it can actually reflect a mechanism for such activationin vivo. Most importantly, regardless of whether nemosis is found in physiological surroundings, it may hold therapeutic potential in situations where a wound healing- or tissue regeneration response is either desired in a controlled fashion or must be suppressed.

Concluding remarks | 55 Acknowledgements

A PhD thesis is written by one person, but contributed to by a whole bunch. Six and a half years at the Haartman Institute have gone by in a flash, but yet been long enough to get to know so many wonderful people who, all in their own ways, contributed to the completion of this thesis. I carried out my work at the Depart- ment of Virology and wish to thank the Head of the Department, Kalle Saksela, for providing the facilities required.

Antti Vaheri, my supervisor: I am humbled by your endless inspiration, optimism, and belief in others (not to mention your total citation and publication count). You have shown me the importance of networking and how simply daring to ask for a favor, kindly and without shame, often gives a positive outcome. It has always been a pleasure to work with you and I cannot think of a single argument we had during the years that did not end in a compromise and satisfied smiles. I honestly think the worst arguments we had concerned the relatively unimportant typographical lay-out of research manuscripts. I am truly grateful to you for the freedom you gave regarding my choice of projects, as well as how to execute them. Lastly, I would like to thank you for the extensive work you did regarding countless MBA school and scholarship recommendations. Reviewers and custos: I would like to thank my custos and undergraduate as well as post-graduate professor Carl Gahmberg for his input in preparing this dissertation. I am also very much indebted to him for introducing me to the fas- cinating world of inflammation and vascular research, which led me first to the Netherlands for a year’s work practice, and ultimately to my PhD project here in Finland. I am also grateful to the reviewers of this thesis, Hannu Larjava and Pirjo Laakkonen, and their valuable critique, which certainly improved this thesis, as well as their permission to proceed with the preparation of this dissertation. Graduate school and thesis follow-up group: I would like to extend my gratitude to the support the Helsinki Graduate School in Biotechnology and Mo- lecular Biology has given me, financially, scientifically, and socially. Without it my PhD experience would have been very different indeed. Pirjo Laakkonen and Ari Ristimäki, my thesis follow-up group members, are sincerely thanked for the com- ments and feed-back received through the years.

56 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Nemosis group: Perttu, the incurable sceptic (realist?), thank you for your valuable input in my work. You have provided answers to many questions, and critique to my thoughts and manuscripts. Thank you also for the support and good times in the lab and office, as well as writing that crucial business school recom- mendation. Kati, you and I are very different and yet very much alike. When I think of the good times I remember wine tasting in New Mexico, shop ‘til you drop in New York, horse carriage rides in Greece, and hyperthermia and ice-cold drinks in the hot tub in the Rockies. I admire your drive and wish you all the best of luck with your new challenges on the other side of the pond. And last but not least, my heartfelt thanks to the technical knowledge vault of our lab, Irina. I am grateful for your assistance, skills, and advice from the very first day I entered the lab. It has been a pleasure working with you - your preciseness and conscientiousness is an inspiration. Zoonosis group and tag-alongs: My sincere thanks goes out to all of you related to the Zoonosis Group and Department of Virology at the Haartman In- stitute. You are too many to be named without forgetting one or two, and so I attempt no such thing. Thank you colleagues of the fourth floor coffee room for great escapes from scientific thought, and thank all of you people for great escapes to Hadanka, or often no further than to a Friday afternoon coffee room filled with people, laughter, and sparkling wine. It has been a pleasure sharing working days with you. Especially thank you Erika for your friendship, optimism, and your constantly good humor, as well as for some interesting movie experiences during the years. And thank you Hao, the outlier in the Nemosis office. We have had some interesting discussions during the years and I have truly enjoyed your company. I wish you all the best for your future aspirations. Marjaleena, Oskari, and Sari, a.k.a. the collaboration formerly known as “plasminogen activation coffee club”, thank you for the conversations, coffee, delicious pastries, and the lab space I have constantly borrowed. Finally I would like to thank the excellent lab technician ladies Leena and Tytti, who have kindly assisted me with everything I ever asked. The Vik-gänget and the reformed Lunch Club: Maria L., Mia, Penny, San- dra, and Tomas. Although we are somewhat dispersed today, we spent years study- ing together both downtown at the old Zoologicum and Botanicum, and in the Biocenter in Viikki. Thank you for the support you gave, whether it was when I had to pick on a smelly dead rat, or plough through Stryer’s Biochemistry, or repeat one

Concluding remarks | 57 whole task during a biochemistry lab course. Thanks also for the long Wednesday gossip lunches in Biomedicum during my first years of PhD work. I have absolutely enjoyed these good times. Friends: I am especially grateful to Maria P. for constant lunch company and never-failing friendship throughout the years. Thank you for your wonderful support and good advice during tough times, as well as giving me so many reasons to laugh and forget about work for a while. You found your calling somewhere else, but I feel our friendship has not changed even with less frequent interaction. I would also like to thank my wonderful friends outside of the Meilahti campus who have provided excellent distraction, entertainment, and support, some all the way from elementary school, others from high school, and several since the past few years. Thank you for your friendship and the good times, and sorry for not keeping in touch as frequently as I should have. Äiti, Pappa, and Nina: Thank you for your support and understanding. I have chosen my own challenges and I know you didn’t always understand why I had to work in the weekends. Now I will be leaving you for a year and again ask for your understanding. We are a busy family, but it feels good to know that no matter where I go or what I do, you will always be there for me. In a world of uncertainty, this is a luxury to hold on to. I hope I can one day return the trust you have shown. Arno: I am grateful for the time you took for listening to my presentation re- hearsals and my general rambles about work and science, as well as for your blood (literally). Thank you also for producing the lay-out and guiding the print of this thesis book. I am forever indebted to you for never protesting when I worked almost every weekend during the last year or when I forced you to help remove sample racks from a broken freezer at 11 p.m. on a Saturday night, for taking all those late-night hails of bad temper I threw upon you in exhaustion, and for believing in me when I did not. You are my soulmate and solid ground.

Espoo, July 2010

Anna Enzerink

58 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation He dug the soil in rows, imposed himself with shovels He asserted into the furrows, I am not random.

The ground replied with aphorisms: a tree-sprout, a nameless weed, words he couldn’t understand.

But obstinate he stated, The land is solid and stamped, watching his foot sink down through stone up to the knee.

From Progressive insanities of a pioneer by Margaret Atwood References

Adams R. H. and Alitalo K. (2007) Molecular regulation of angiogenesis and lymphangiogenesis. Nat Rev Mol Cell Biol 8, 464-78. Aghajanian A., Wittchen E. S., Allingham M. J., Garrett T. A. and Burridge K. (2008) Endothelial cell junctions and the regulation of vascular permeability and leukocyte transmigration. J Thromb Haemost 6, 1453-60. Allport J. R., Muller W. A. and Luscinskas F. W. (2000) Monocytes induce reversible focal changes in vascular endothelial cadherin complex during transendothelial migration under flow. J Cell Biol 148, 203-16. Badiavas E. V., Abedi M., Butmarc J., Falanga V. and Quesenberry P. (2003) Participation of bone marrow derived cells in cutaneous wound healing. J Cell Physiol 196, 245-50. Baglole C. J., Ray D. M., Bernstein S. H., Feldon S. E., Smith T. J., Sime P. J. and Phipps R. P. (2006) More than structural cells, fibroblasts create and orchestrate the tumor microenvironment.Immunol Invest 35, 297-325. Balkwill F., Charles K. A. and Mantovani A. (2005) Smoldering and polarized inflammation in the initiation and promotion of malignant disease. Cancer Cell 7, 211-217. Balkwill F. and Mantovani A. (2001) Inflammation and cancer: back to Virchow? Lancet 357, 539-45. Bao P., Kodra A., Tomic-Canic M., Golinko M. S., Ehrlich H. P. and Brem H. (2009) The role of vascular endothelial growth factor in wound healing. J Surg Res 153, 347-58. Bazzoni G. (2006) Endothelial tight junctions: permeable barriers of the vessel wall. Thromb Haemost 95, 36-42. Berthod F., Germain L., Tremblay N. and Auger F. A. (2006) Extracellular matrix deposition by fibroblasts is necessary to promote capillary-like tube formation in vitro. J Cell Physiol 207, 491-8. Birchmeier C., Birchmeier W., Gherardi E. and Vande Woude G. F. (2003) Met, metastasis, motility and more. Nat Rev Mol Cell Biol 4, 915-925. Bizik J., Kankuri E., Ristimaki A., Taieb A., Vapaatalo H., Lubitz W. and Vaheri A. (2004) Cell-cell contacts trigger programmed necrosis and induce cyclooxygenase-2 expression. Cell Death Differ 11, 183-95. Boccaccio C., Sabatino G., Medico E., Girolami F., Follenzi A., Reato G., Sottile A., Naldini L. and Comoglio P. M. (2005) The MET oncogene drives a genetic programme linking cancer to haemostasis. Nature 434, 396- 400. Boukamp P., Petrussevska R. T., Bretikreutz D., Hornung J., Markham A. and Fusenig N. E. (1988) Normal keratiniza- tion in a spontaneously immortalized aneuploid human keratinocyte cell line. J Cell Biol 106, 761-771. Braiman-Wiksman L., Solomonik I., Spira R. and Tennenbaum T. (2007) Novel insights into wound healing sequence of events. Toxicol Pathol 35, 767-79. Broughton G., 2nd, Janis J. E. and Attinger C. E. (2006) A brief history of wound care. Plast Reconstr Surg 117, 6S- 11S. Brouty-Boyé D., Pottin-Clémenceau C., Doucet C., Jasmin C. and Azzarone B. (2000) Chemokines and CD40 expression in human fibroblasts. Eur J Immunol 30, 914-919. Brown L. F., Yeo K. T., Berse B., Yeo T. K., Senger D. R., Dvorak H. F. and van de Water L. (1992) Expression of vascular permeability factor (vascular endothelial growth factor) by epidermal keratinocytes during wound healing. J Exp Med 176, 1375-9. Buckley C. D., Pilling D., Lord J. M., Akbar A. N., Scheel-Toellner D. and Salmon M. (2001) Fibroblasts regulate the switch from acute resolving to chronic persistent inflammation. Trends Immunol 22, 199-204. Bussolino F., Di Renzo M. F., Ziche M., Bocchietto E., Olivero M., Naldini L., Gaudino G., Tamagnone L., Coffer A. and Comoglio P. M. (1992) Hepatocyte growth factor is a potent angiogenic factor which stimulates endothelial cell motility and growth. J Cell Biol 119, 629-41. Carman C. V. and Springer T. A. (2004) A transmigratory cup in leukocyte diapedesis both through individual vascular endothelial cells and between them. J Cell Biol 167, 377-88. Carmeliet P., Lampugnani M. G., Moons L., Breviario F., Compernolle V., Bono F., Balconi G., Spagnuolo R., Oost- huyse B., Dewerchin M., Zanetti A., Angellilo A., Mattot V., Nuyens D., Lutgens E., Clotman F., de Ruiter M. C., Gittenberger-de Groot A., Poelmann R., Lupu F., Herbert J. M., Collen D. and Dejana E. (1999) Targeted deficiency or cytosolic truncation of the VE-cadherin gene in mice impairs VEGF-mediated endothelial survival and angiogenesis. Cell 98, 147-57. Castello G., Scala S., Palmieri G., Curley S. A. and Izzo F. (2010) HCV-related hepatocellular carcinoma: From chronic inflammation to cancer. Clin Immunol 134, 237-50. Caveda L., Martin-Padura I., Navarro P., Breviario F., Corada M., Gulino D., Lampugnani M. G. and Dejana E. (1996) Inhibition of cultured cell growth by vascular endothelial cadherin (cadherin-5/VE-cadherin). J Clin Invest 98, 886-93. Chang H. Y., Chi J. T., Dudoit S., Bondre C., van de Rijn M., Botstein D. and Brown P. O. (2002) Diversity, topographic

60 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation differentiation, and positional memory in human fibroblasts. Proc Natl Acad Sci U S A 99, 12877-82. Chang S. H., Liu C. H., Conway R., Han D. K., Nithipatikom K., Trifan O. C., Lane T. F. and Hla T. (2004) Role of prostaglandin E2-dependent angiogenic switch in cyclooxygenase 2-induced breast cancer progression. Proc Natl Acad Sci U S A 101, 591-6. Chavakis E., Choi E. Y. and Chavakis T. (2009) Novel aspects in the regulation of the leukocyte adhesion cascade. Thromb Haemost 102, 191-7. Chemin I. and Zoulim F. (2009) Hepatitis B virus induced hepatocellular carcinoma. Cancer Lett 286, 52-9. Chiarugi P. and Giannoni E. (2008) Anoikis: a necessary death program for anchorage-dependent cells. Biochem Pharmacol 76, 1352-64. Chida M. and Voelkel N. F. (1996) Effects of acute and chronic hypoxia on rat lung cyclooxygenase. Am J Physiol 270, L872-8. Chmielowiec J., Borowiak M., Morkel M., Stradal T., Munz B., Werner S., Wehland J., Birchmeier C. and Birchmeier W. (2007) c-Met is essential for wound healing in the skin. J Cell Biol 177, 151-162. Cohnheim J. (1867) Ueber Entzündung und Eiterung. Virchows Arch 40, 1-79. Colotta F., Allavena P., Sica A., Garlanda C. and Mantovani A. (2009) Cancer-related inflammation, the seventh hallmark of cancer: links to genetic instability. Carcinogenesis 30, 1073-81. Conway K., Price P., Harding K. G. and Jiang W. G. (2006) The molecular and clinical impact of hepatocyte growth factor, its receptor, activators, and inhibitors in wound healing. Wound Repair Regen 14, 2-10. Corada M., Mariotti M., Thurston G., Smith K., Kunkel R., Brockhaus M., Lampugnani M. G., Martin-Padura I., Stop- pacciaro A., Ruco L., McDonald D. M., Ward P. A. and Dejana E. (1999) Vascular endothelial-cadherin is an important determinant of microvascular integrity in vivo. Proc Natl Acad Sci U S A 96, 9815-20. Crofford L. J., Wilder R. L., Ristimaki A. P., Sano H., Remmers E. F., Epps H. R. and Hla T. (1994) Cyclooxygenase-1 and -2 expression in rheumatoid synovial tissues. Effects of interleukin-1 beta, phorbol ester, and corti- costeroids. J Clin Invest 93, 1095-101. Crunkhorn P. and Willis A. L. (1971) Cutaneous reactions to intradermal prostaglandins. Br J Pharmacol 41, 49-56. Cukierman E., Pankov R., Stevens D. R. and Yamada K. M. (2001) Taking cell-matrix adhesions to the third dimen- sion. Science 294, 1708-12. Cukierman E., Pankov R. and Yamada K. M. (2002) Cell interactions with three-dimensional matrices. Curr Opin Cell Biol 14, 633-9. De Wever O., Demetter P., Mareel M. and Bracke M. (2008) Stromal myofibroblasts are drivers of invasive cancer growth. Int J Cancer 123, 2229-38. Dejana E., Orsenigo F. and Lampugnani M. G. (2008) The role of adherens junctions and VE-cadherin in the control of vascular permeability. J Cell Sci 121, 2115-22. Desmouliere A., Chaponnier C. and Gabbiani G. (2005) Tissue repair, contraction, and the myofibroblast.Wound Repair Regen 13, 7-12. Dong G., Chen Z., Li Z. Y., Yeh N. T., Bancroft C. C. and Van Waes C. (2001) Hepatocyte growth factor/scatter factor-induced activation of MEK and PI3K signal pathways contributes to expression of proangiogenic cytokines interleukin-8 and vascular endothelial growth factor in head and neck squamous cell carcinoma. Cancer Res 61, 5911-8. Dustin M. L., Rothlein R., Bhan A. K., Dinarello C. A. and Springer T. A. (1986) Induction by IL 1 and interferon- gamma: tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-1). J Im- munol 137, 245-54. Ehrlich H. P. and Diez T. (2003) Role for gap junctional intercellular communications in wound repair. Wound Repair Regen 11, 481-489. Eming S. A., Brachvogel B., Odorisio T. and Koch M. (2007) Regulation of angiogenesis: wound healing as a model. Prog Histochem Cytochem 42, 115-70. Engelhardt E., Toksoy A., Goebeler M., Debus S., Brocker E. B. and Gillitzer R. (1998) Chemokines IL-8, GROalpha, MCP-1, IP-10, and Mig are sequentially and differentially expressed during phase-specific infiltration of leukocyte subsets in human wound healing. Am J Pathol 153, 1849-60. Erez N., Truitt M., Olson P. and Hanahan D. (2010) Cancer-Associated Fibroblasts Are Activated in Incipient Neo- plasia to Orchestrate Tumor-Promoting Inflammation in an NF-kappaB-Dependent Manner.Cancer Cell 17, 135-147. Estes J. M., Vande Berg J. S., Adzick N. S., MacGillivray T. E., Desmouliere A. and Gabbiani G. (1994) Phenotypic and functional features of myofibroblasts in sheep fetal wounds. Differentiation 56, 173-81. Eyden B. (2008) The myofibroblast: phenotypic characterization as a prerequisite to understanding its functions in translational medicine. J Cell Mol Med 12, 22-37. Eyden B. (2001) The myofibroblast: an assessment of controversial issues and a definition useful in diagnosis and research. Ultrastruct Pathol 25, 39-50.

References | 61 Fan S., Gao M., Meng Q., Laterra J. J., Symons M. H., Coniglio S., Pestell R. G., Goldberg I. D. and Rosen E. M. (2005) Role of NF-kB signaling in hepatocyte growth factor/scatter factor-mediated cell protection. Oncogene 24, 1749-1766. Fanning A. S. and Anderson J. M. (2009) Zonula occludens-1 and -2 are cytosolic scaffolds that regulate the assembly of cellular junctions. Ann N Y Acad Sci 1165, 113-20. Farquhar M. G. and Palade G. E. (1963) Junctional complexes in various epithelia. J Cell Biol 17, 375-412. Feng D., Nagy J. A., Pyne K., Dvorak H. F. and Dvorak A. M. (1998) Neutrophils emigrate from venules by a transendothelial cell pathway in response to FMLP. J Exp Med 187, 903-15. Ferrara N., Carver-Moore K., Chen H., Dowd M., Lu L., O’Shea K. S., Powell-Braxton L., Hillan K. J. and Moore M. W. (1996) Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439-42. Ferrara N., Gerber H. P. and LeCouter J. (2003) The biology of VEGF and its receptors. Nat Med 9, 669-76. Findley M. K. and Koval M. (2009) Regulation and roles for claudin-family tight junction proteins. IUBMB Life 61, 431-7. Flavell S. J., Hou T. Z., Lax S., Filer A. D., Salmon M. and Buckley C. D. (2008) Fibroblasts as novel therapeutic targets in chronic inflammation. Br J Pharmacol 153 Suppl 1, S241-6. Folkman J., Merler E., Abernathy C. and Williams G. (1971) Isolation of a tumor factor responsible for angiogenesis. J Exp Med 133, 275-88. Freter R. R., Alberta J. A., Hwang G. Y., Wrentmore A. L. and Stiles C. D. (1996) Platelet-derived growth factor induction of the immediate-early gene MCP-1 is mediated by NF-kappaB and a 90-kDa phosphoprotein coactivator. J Biol Chem 271, 17417-24. Fukumura D., Xavier R., Sugiura T., Chen Y., Park E. C., Lu N., Selig M., Nielsen G., Taksir T., Jain R. K. and Seed B. (1998) Tumor induction of VEGF promoter activity in stromal cells. Cell 94, 715-25. Gabbiani G. (2004). The evolution of the myofibroblast concept: a key cell for wound healing and fibrotic diseases. G Gerontol 52, 280-282 Gabbiani G., Chaponnier C. and Huttner I. (1978) Cytoplasmic filaments and gap junctions in epithelial cells and myofibroblasts during wound healing. J Cell Biol 76, 561-568. Gabbiani G., Ryan G. B. and Majne G. (1971) Presence of modified fibroblasts in granulation tissue and their possible role in wound contraction. Experientia 27, 549-50. Gahmberg C. G., Fagerholm S. C., Nurmi S. M., Chavakis T., Marchesan S. and Gronholm M. (2009) Regulation of integrin activity and signalling. Biochim Biophys Acta 1790, 431-44. Gahmberg C. G., Tolvanen M. and Kotovuori P. (1997) Leukocyte adhesion--structure and function of human leukocyte beta2-integrins and their cellular ligands. Eur J Biochem 245, 215-32. Garrido-Urbani S., Bradfield P. F., Lee B. P. and Imhof B. A. (2008) Vascular and epithelial junctions: a barrier for leucocyte migration. Biochem Soc Trans 36, 203-11. Gavard J. and Gutkind J. S. (2006) VEGF controls endothelial-cell permeability by promoting the beta-arrestin- dependent endocytosis of VE-cadherin. Nat Cell Biol 8, 1223-34. Gerber P. A., Hippe A., Buhren B. A., Muller A. and Homey B. (2009) Chemokines in tumor-associated angiogenesis. Biol Chem 390, 1213-23. Gerhardt H., Golding M., Fruttiger M., Ruhrberg C., Lundkvist A., Abramsson A., Jeltsch M., Mitchell C., Alitalo K., Shima D. and Betsholtz C. (2003) VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 161, 1163-77. Gerritsen M. E., Tomlinson J. E., Zlot C., Ziman M. and Hwang S. (2003) Using gene expression profiling to iden- tify the molecular basis of the synergistic actions of hepatocyte growth factor and vascular endothelial growth factor in human endothelial cells. Br J Pharmacol 140, 595-610. Gill S. E. and Parks W. C. (2008) Metalloproteinases and their inhibitors: regulators of wound healing. Int J Biochem Cell Biol 40, 1334-47. Gilmore T. D. (2006) Introduction to NF-kappaB: players, pathways, perspectives. Oncogene 25, 6680-4. González-Cuadrado S., Bustos C., Ruiz-Ortega M., Ortiz A., Guijarro C., Plaza J. J. and Egido J. (1996) Expression of leucocyte chemoattractants by interstitial renal fibroblasts: up-regulation by drugs associated with interstitial fibrosis. Clin Exp Immunol 106, 518-522. Granger D. N. and Kubes P. (1994) The microcirculation and inflammation: modulation of leukocyte-endothelial cell adhesion. J Leukoc Biol 55, 662-75. Greten F. R., Eckmann L., Greten T. F., Park J. M., Li Z. W., Egan L. J., Kagnoff M. F. and Karin M. (2004) IKK-beta links inflammation and tumorigenesis in a mouse model of colitis-associated cancer.Cell 118, 285-296. Grunewald M., Avraham I., Dor Y., Bachar-Lustig E., Itin A., Jung S., Chimenti S., Landsman L., Abramovitch R. and Keshet E. (2006) VEGF-induced adult neovascularization: recruitment, retention, and role of accessory cells. Cell 124, 175-89.

62 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Gurtner G. C., Werner S., Barrandon Y. and Longaker M. T. (2008) Wound repair and regeneration. Nature 453, 314-21. Han S., Sidell N., Roser-Page S. and Roman J. (2004) Fibronectin stimulates human lung carcinoma cell growth by inducing cyclooxygenase-2 (COX-2) expression. Int J Cancer 111, 322-31. Hanahan D. and Weinberg R. A. (2000) The hallmarks of cancer. Cell 100, 57-70. Harding K. G., Moore K. and Phillips T. J. (2005) Wound chronicity and fibroblast senescence--implications for treatment. Int Wound J 2, 364-8. Hattori R., Hamilton K. K., Fugate R. D., McEver R. P. and Sims P. J. (1989) Stimulated secretion of endothelial von Willebrand factor is accompanied by rapid redistribution to the cell surface of the intracellular granule membrane protein GMP-140. J Biol Chem 264, 7768-71. Hayden M. S., West A. P. and Ghosh S. (2006) NF-kappaB and the immune response. Oncogene 25, 6758-80. Hellstrom M., Phng L. K., Hofmann J. J., Wallgard E., Coultas L., Lindblom P., Alva J., Nilsson A. K., Karlsson L., Gaiano N., Yoon K., Rossant J., Iruela-Arispe M. L., Kalen M., Gerhardt H. and Betsholtz C. (2007) Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445, 776-80. Hinz B. and Gabbiani G. (2003) Cell-matrix and cell-cell contacts of myofibroblasts: role in connective tissue remodeling. Thromb Haemost 90, 993-1002. Hinz B. (2007a) Formation and function of the myofibroblast during tissue repair. J Invest Dermatol 127, 526-37. Hinz B., Phan S. H., Thannickal V. J., Galli A., Bochaton-Piallat M. L. and Gabbiani G. (2007b) The myofibroblast, one function, multiple origins. Am J Pathol 170, 1807-1816. Hogaboam C. M., Steinhauser M. L., Chensue S. W. and Kunkel S. L. (1998) Novel roles for chemokines and fibroblasts in interstitial fibrosis. Kidney Int 54, 2152-2159. Hughes C. C. (2008) Endothelial-stromal interactions in angiogenesis. Curr Opin Hematol 15, 204-9. Itoh M., Furuse M., Morita K., Kubota K., Saitou M. and Tsukita S. (1999) Direct binding of three tight junction-associated MAGUKs, ZO-1, ZO-2, and ZO-3, with the COOH termini of claudins. J Cell Biol 147, 1351-63. Jiang W., Puntis M. C., Nakamura T. and Hallett M. B. (1992) Neutrophil priming by hepatocyte growth factor, a novel cytokine. Immunology 77, 147-9. Kajiya K., Hirakawa S., Ma B., Drinnenberg I. and Detmar M. (2005) Hepatocyte growth factor promotes lymphatic vessel formation and function. EMBO J 24, 2885-95. Kalluri R. and Zeisberg M. (2006) Fibroblasts in cancer. Nat Rev Cancer 6, 392-401. Kankuri E., Babusikova O., Hlubinova K., Salmenperä P., Boccaccio C., Lubitz W., Harjula A. and Bizik J. (2008) Fibroblast nemosis arrests growth and induces differentiation of human leukemia cells. Int J Cancer 122, 1243-52. Kankuri E., Cholujova D., Comajova M., Vaheri A. and Bizik J. (2005) Induction of HGF/SF by cell-cell contacts in fibroblasts directly promotes tumor invasiveness. Cancer Res 65, 9914-9922. Karin M. (2006) Nuclear factor-kB in cancer development and progression. Nature 441, 431-436. Katsuno T., Umeda K., Matsui T., Hata M., Tamura A., Itoh M., Takeuchi K., Fujimori T., Nabeshima Y., Noda T. and Tsukita S. (2008) Deficiency of zonula occludens-1 causes embryonic lethal phenotype associated with defected yolk sac angiogenesis and apoptosis of embryonic cells. Mol Biol Cell 19, 2465-75. Keeley E. C., Mehrad B. and Strieter R. M. (2008) Chemokines as mediators of neovascularization. Arterioscler Thromb Vasc Biol 28, 1928-36. Kiriakidis S., Andreakos E., Monaco C., Foxwell B., Feldmann M. and Paleolog E. (2003) VEGF expression in human macrophages is NF-kappaB-dependent: studies using adenoviruses expressing the endogenous NF-kap- paB inhibitor IkappaBalpha and a kinase-defective form of the IkappaB kinase 2. J Cell Sci 116, 665-74. Kisseleva T., Song L., Vorontchikhina M., Feirt N., Kitajewski J. and Schindler C. (2006) NF-kappaB regulation of endothelial cell function during LPS-induced toxemia and cancer. J Clin Invest 116, 2955-63. Klagsbrun M., Knighton D. and Folkman J. (1976) Tumor angiogenesis activity in cells grown in tissue culture. Cancer Res 36, 110-4. Kolachala V. L., Bajaj R., Wang L., Yan Y., Ritzenthaler J. D., Gewirtz A. T., Roman J., Merlin D. and Sitaraman S. V. (2007) Epithelial-derived fibronectin expression, signaling, and function in intestinal inflammation. J Biol Chem 282, 32965-73. Kumar I., Staton C. A., Cross S. S., Reed M. W. and Brown N. J. (2009) Angiogenesis, vascular endothelial growth factor and its receptors in human surgical wounds. Br J Surg 96, 1484-91. Lambeng N., Wallez Y., Rampon C., Cand F., Christe G., Gulino-Debrac D., Vilgrain I. and Huber P. (2005) Vascular endothelial-cadherin tyrosine phosphorylation in angiogenic and quiescent adult tissues. Circ Res 96, 384-91. Lampugnani M. G., Resnati M., Raiteri M., Pigott R., Pisacane A., Houen G., Ruco L. P. and Dejana E. (1992) A novel endothelial-specific membrane protein is a marker of cell-cell contacts. J Cell Biol 118, 1511-22. Lang F. J. (2001) Role of endothelium in the production of polyblasts (mononuclear wandering cells) in inflamma-

References | 63 tion. 1926. Arch Pathol Lab Med 125, 44-66. Langenkamp E. and Molema G. (2009) Microvascular endothelial cell heterogeneity: general concepts and phar- macological consequences for anti-angiogenic therapy of cancer. Cell Tissue Res 335, 205-22. Larjava H., Salo T., Haapasalmi K., Kramer R. H. and Heino J. (1993) Expression of integrins and basement membrane components by wound keratinocytes. J Clin Invest 92, 1425-35. Lawson C. and Wolf S. (2009) ICAM-1 signaling in endothelial cells. Pharmacol Rep 61, 22-32. Levy B. D., Clish C. B., Schmidt B., Gronert K. and Serhan C. N. (2001) Lipid mediator class switching during acute inflammation: signals in resolution. Nat Immunol 2, 612-9. Li H., Bauzon D. E., Xu X., Tschesche H., Cao J. and Sang Q. A. (1998) Immunological characterization of cell-surface and soluble forms of membrane type 1 matrix metalloproteinase in human breast cancer cells and in fibroblasts. Mol Carcinog 22, 84-94. Li J., Chen J. and Kirsner R. (2007) Pathophysiology of acute wound healing. Clin Dermatol 25, 9-18. Li L., Dragulev B., Zigrino P., Mauch C. and Fox J. W. (2009) The invasive potential of human melanoma cell lines correlates with their ability to alter fibroblast gene expressionin vitro and the stromal microenvironment in vivo. Int J Cancer 125, 1796-804. Li Q. (2002) NF-kB regulation in the immune system. Nat Rev Immunol 2, 725-735. Liu X. H., Kirschenbaum A., Yao S., Stearns M. E., Holland J. F., Claffey K. and Levine A. C. (1999) Upregulation of vascular endothelial growth factor by cobalt chloride-simulated hypoxia is mediated by persistent induction of cyclooxygenase-2 in a metastatic human prostate cancer cell line. Clin Exp Metastasis 17, 687-94. Lo D., Feng L., Li L., Carson M. J., Crowley M., Pauza M., Nguyen A. and Reilly C. R. (1999) Integrating innate and adaptive immunity in the whole animal. Immunol Rev 169, 225-39. Lukacs N. W., Chensue S. W., Smith R. E., Strieter R. M., Warmington K., Wilke C. and Kunkel S. L. (1994) Production of monocyte chemoattractant protein-1 and macrophage inflammatory protein-1alpha by inflammatory granuloma fibroblasts. Am J Pathol 144, 711-718. Maiellaro K. and Taylor W. R. (2007) The role of the adventitia in vascular inflammation.Cardiovasc Res 75, 640- 8. Maisi P., Prikk K., Sepper R., Pirila E., Salo T., Hietanen J. and Sorsa T. (2002) Soluble membrane-type 1 matrix metalloproteinase (MT1-MMP) and gelatinase A (MMP-2) in induced sputum and bronchoalveolar lavage fluid of human bronchial asthma and bronchiectasis. APMIS 110, 771-82. Mantovani A., Allavena P., Sica A. and Balkwill F. (2008) Cancer-related inflammation. Nature 454, 436-44. Mars W. M., Zarnegar R. and Michalopoulos G. K. (1993) Activation of hepatocyte growth factor by the plasminogen activators uPA and tPA. Am J Pathol 143, 949-58. Martin T., Cardarelli P. M., Parry G. C., Felts K. A. and Cobb R. R. (1997) Cytokine induction of monocyte chemoattractant protein-1 gene expression in human endothelial cells depends on the cooperative action of NF- kappa B and AP-1. Eur J Immunol 27, 1091-7. Matsunaga T., Gohda E., Takebe T., Wu Y. L., Iwao M., Kataoka H. and Yamamoto I. (1994) Expression of hepatocyte growth factor is up-regulated through activation of a cAMP-mediated pathway. Exp Cell Res 210, 326- 35. Mauviel A., Reitamo S., Remitz A., Lapiere J. C., Ceska M., Baggiolini M., Walz A., Evans C. H. and Uitto J. (1992) Leu- koregulin, a T cell-derived cytokine, induces IL-8 gene expression and secretion in human skin fibroblasts. Demonstration and secretion in human skin fibroblasts. Demonstration of enhanced NF-kappa B binding and NF-kappa B-driven promoter activity. J Immunol 149, 2969-76. McAnulty R. J. (2007) Fibroblasts and myofibroblasts: their source, function and role in disease.Int J Biochem Cell Biol 39, 666-71. Medzhitov R. (2008) Origin and physiological roles of inflammation. Nature 454, 428-35. Micke P. and Östman A. (2004) Tumour-stroma interaction: cancer-associated fibroblasts as novel targets in anti- cancer therapy? Lung Cancer 45, S163-S175. Mine S., Tanaka Y., Suematu M., Aso M., Fujisaki T., Yamada S. and Eto S. (1998) Hepatocyte growth factor is a potent trigger of neutrophil adhesion through rapid activation of lymphocyte function-associated antigen-1. Lab Invest 78, 1395-404. Miyazawa K., Shimomura T. and Kitamura N. (1996) Activation of hepatocyte growth factor in the injured tissues is mediated by hepatocyte growth factor activator. J Biol Chem 271, 3615-8. Moghul A., Lin L., Beedle A., Kanbour-Shakir A., DeFrances M. C., Liu Y. and Zarnegar R. (1994) Modulation of c-MET proto-oncogene (HGF receptor) mRNA abundance by cytokines and hormones: evidence for rapid decay of the 8 kb c-MET transcript. Oncogene 9, 2045-52. Monnier Y., Zaric J. and Ruegg C. (2005) Inhibition of angiogenesis by non-steroidal anti-inflammatory drugs: from the bench to the bedside and back. Curr Drug Targets Inflamm Allergy 4, 31-8. Montesano R., Pepper M. S. and Orci L. (1993) Paracrine induction of angiogenesis in vitro by Swiss 3T3 fibroblasts.

64 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation J Cell Sci 105 ( Pt 4), 1013-24. Morita K., Sasaki H., Furuse M. and Tsukita S. (1999) Endothelial claudin: claudin-5/TMVCF constitutes tight junction strands in endothelial cells. J Cell Biol 147, 185-94. Mornex J. F., Leroux C., Greenland T. and Ecochard D. (1994) From granuloma to fibrosis in interstitital lung diseases: molecular and cellular interactions. Eur Respir J 7, 779-785. Mrowietz U., Schwenk U., Maune S., Bartels J., Küpper M., Fichtner I., Schröder J. M. and Schadendorf D. (1999) The chemokine RANTES is secreted by human melanoma cells and is associated with enhanced tumour formation in nude mice. Br J Cancer 79, 1025-1031. Mueller M. M. and Fusenig N. E. (2002) Tumor-stroma interactions directing phenotype and progression of epithelial skin tumor cells. Differentiation 70, 486-497. Muller W. A. (2009) Mechanisms of transendothelial migration of leukocytes. Circ Res 105, 223-30. Nagy J. A., Benjamin L., Zeng H., Dvorak A. M. and Dvorak H. F. (2008) Vascular permeability, vascular hyperperme- ability and angiogenesis. Angiogenesis 11, 109-19. Nakamura T., Teramoto H. and Ichihara A. (1986) Purification and characterization of a growth factor from rat platelets for mature parenchymal hepatocytes in primary cultures. Proc Natl Acad Sci U S A 83, 6489- 93. Nakamura Y., Morishita R., Higaki J., Kida I., Aoki M., Moriguchi A., Yamada K., Hayashi S., Yo Y., Matsumoto K., Na- kamura T. and Ogihara T. (1995) Expression of local hepatocyte growth factor system in vascular tissues. Biochem Biophys Res Commun 215, 483-8. Nakatsu M. N., Sainson R. C., Aoto J. N., Taylor K. L., Aitkenhead M., Perez-del-Pulgar S., Carpenter P. M. and Hughes C. C. (2003) Angiogenic sprouting and capillary lumen formation modeled by human umbilical vein endothelial cells (HUVEC) in fibrin gels: the role of fibroblasts and Angiopoietin-1.Microvasc Res 66, 102-12. Newton R., Kuitert L. M., Bergmann M., Adcock I. M. and Barnes P. J. (1997) Evidence for involvement of NF-kappaB in the transcriptional control of COX-2 gene expression by IL-1beta. Biochem Biophys Res Commun 237, 28-32. Nissen N. N., Polverini P. J., Koch A. E., Volin M. V., Gamelli R. L. and DiPietro L. A. (1998) Vascular endothelial growth factor mediates angiogenic activity during the proliferative phase of wound healing. Am J Pathol 152, 1445-52. Nitta T., Hata M., Gotoh S., Seo Y., Sasaki H., Hashimoto N., Furuse M. and Tsukita S. (2003) Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice. J Cell Biol 161, 653-60. Nor J. E., Christensen J., Liu J., Peters M., Mooney D. J., Strieter R. M. and Polverini P. J. (2001) Up-Regulation of Bcl- 2 in microvascular endothelial cells enhances intratumoral angiogenesis and accelerates tumor growth. Cancer Res 61, 2183-8. Nyqvist D., Giampietro C. and Dejana E. (2008) Deciphering the functional role of endothelial junctions by using in vivo models. EMBO Rep 9, 742-7. O’Byrne K. J. and Dalgleish A. G. (2001) Chronic immune activation and inflammation as the cause of malignancy. Br J Cancer 85, 473-483. Olumi A. F., Grossfeld G. D., Hayward S. W., Carroll P. R., Tlsty T. D. and Cunha G. R. (1999) Carcinoma-associated fibroblasts direct tumor progression of initiated human prostatic epithelium.Cancer Res 59, 5002-11. Orimo A., Gupta P. B., Sgroi D. C., Arenzana-Seisdedos F., Delaunay T., Naeem R., Carey V. J., Richardson A. L. and Weinberg R. A. (2005) Stromal fibroblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through elevated SDF-1/CXCL12 secretion. Cell 121, 335-48. Orimo A. and Weinberg R. A. (2006) Stromal fibroblasts in cancer: a novel tumor-promoting cell type.Cell Cycle 5, 1597-601. Øynebråten I., Bakke O., Brandtzaeg P., Johansen F. E. and Haraldsen G. (2004) Rapid chemokine secretion from endothelial cells originates from 2 distinct compartments. Blood 104, 314-20. Øynebråten I., Barois N., Hagelsteen K., Johansen F. E., Bakke O. and Haraldsen G. (2005) Characterization of a novel chemokine-containing storage granule in endothelial cells: evidence for preferential exocytosis mediated by protein kinase A and diacylglycerol. J Immonol 175, 5358-69. Ozawa M., Baribault H. and Kemler R. (1989) The cytoplasmic domain of the cell adhesion molecule uvomorulin associates with three independent proteins structurally related in different species. EMBO J 8, 1711-7. Pai J. T. and Ruoslahti E. (2005) Identification of endothelial genes up-regulatedin vivo. Gene 347, 21-33. Park M., Dean M., Cooper C. S., Schmidt M., O’Brien S. J., Blair D. G. and Vande Woude G. F. (1986) Mechanism of met oncogene activation. Cell 45, 895-904. Patarroyo M., Clark E. A., Prieto J., Kantor C. and Gahmberg C. G. (1987) Identification of a novel adhesion molecule in human leukocytes by monoclonal antibody LB-2. FEBS Lett 210, 127-31. Peura M., Bizik J., Salmenperä P., Noro A., Korhonen M., Patila T., Vento A., Vaheri A., Alitalo R., Vuola J., Harjula A.

References | 65 and Kankuri E. (2009) Bone marrow mesenchymal stem cells undergo nemosis and induce keratinocyte wound healing utilizing the HGF/c-Met/PI3K pathway. Wound Repair Regen 17, 569-77. Pierce G. B., Pantazis C. G., Caldwell J. E. and Wells R. S. (1982) Specificity of the control of tumor formation by the blastocyst. Cancer Res 42, 1082-7. Põld M., Zhu L. X., Sharma S., Burdick M. D., Lin Y., Lee P. P. N., Põld A., Luo J., Krysan K., Dohadwala M., Mao J. T., Batra R. K., Strieter R. M. and Dubinett S. M. (2004) Cyclooxygenase-2-dependent expression of angiogenic CXC chemokines ENA-78 / CXC ligand (CXCL) 5 and interleukin-8 / CXCL8 in human non-small cell lung cancer. Cancer Res 64, 1853-1860. Powell D. W., Mifflin R. C., Valentich J. D., Crowe S. E., Saada J. I. and West A. B. (1999) Myofibroblasts. I. Paracrine cells important in health and disease. Am J Physiol 277, C1-9. Rahman S., Patel Y., Murray J., Patel K. V., Sumathipala R., Sobel M. and Wijelath E. S. (2005) Novel hepatocyte growth factor (HGF) binding domains on fibronectin and vitronectin coordinate a distinct and amplified Met-integrin induced signalling pathway in endothelial cells. BMC Cell Biol 6, 8. Räsänen K., Salmenperä P., Baumann M., Virtanen I. and Vaheri A. (2008) Nemosis of fibroblasts is inhibited by benign HaCaT keratinocytes but promoted by malignant HaCaT cells. Mol Oncol 2, 340-8. Räsänen K. and Vaheri A. (2010) Proliferation and motility of HaCaT keratinocyte derivatives is enhanced by fibroblast nemosis. Exp Cell Res, in press. doi:10.1016/j.yexcr.2010.01.020. Räsänen K., Virtanen I., Salmenperä P., Grenman R. and Vaheri A. (2009) Differences in the nemosis response of normal and cancer-associated fibroblasts from patients with oral squamous cell carcinoma. PLoS One 4, e6879. Rasband W. S. (1997-2009) ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/. Rechardt O., Elomaa O., Vaalamo M., Paakkonen K., Jahkola T., Hook-Nikanne J., Hembry R. M., Hakkinen L., Kere J. and Saarialho-Kere U. (2000) Stromelysin-2 is upregulated during normal wound repair and is induced by cytokines. J Invest Dermatol 115, 778-87. Rong S., Segal S., Anver M., Resau J. H. and Vande Woude G. F. (1994) Invasiveness and metastasis of NIH 3T3 cells induced by Met-hepatocyte growth factor/scatter factor autocrine stimulation. Proc Natl Acad Sci U S A 91, 4731-5. Roos D. and de Boer M. (1986) Purification and cryopreservation of phagocytes from human blood.Methods Enzymol 132, 225-243. Rossi D. and Zlotnik A. (2000) The biology of chemokines and their receptors. Annu Rev Immunol 18, 217-242. Rot A. and von Andrian U. H. (2004) Chemokines in innate and adaptive host defense: basic chemokinese grammar for immune cells. Annu Rev Immunol 22, 891-928. Ruhrberg C., Gerhardt H., Golding M., Watson R., Ioannidou S., Fujisawa H., Betsholtz C. and Shima D. T. (2002) Spatially restricted patterning cues provided by heparin-binding VEGF-A control blood vessel branching morphogenesis. Genes Dev 16, 2684-98. Sainson R. C., Aoto J., Nakatsu M. N., Holderfield M., Conn E., Koller E. and Hughes C. C. (2005) Cell-autonomous notch signaling regulates endothelial cell branching and proliferation during vascular tubulogenesis. FASEB J 19, 1027-9. Salcedo R., Ponce M. L., Young H. A., Wasserman K., Ward J. M., Kleinman H. K., Oppenheim J. J. and Murphy W. J. (2000) Human endothelial cells express CCR2 and respond to MCP-1: direct role of MCP-1 in angiogenesis and tumor progression. Blood 96, 34-40. Salmenperä P., Kankuri E., Bizik J., Siren V., Virtanen I., Takahashi S., Leiss M., Fassler R. and Vaheri A. (2008) Forma- tion and activation of fibroblast spheroids depend on fibronectin-integrin interaction. Exp Cell Res 314, 3444-52. Sappino A. P., Skalli O., Jackson B., Schurch W. and Gabbiani G. (1988) Smooth-muscle differentiation in stromal cells of malignant and non-malignant breast tissues. Int J Cancer 41, 707-12. Sato T. N. (2003) Emerging concept in angiogenesis: specification of arterial and venous endothelial cells. Br J Pharmacol 140, 611-3. Scatena M., Almeida M., Chaisson M. L., Fausto N., Nicosia R. F. and Giachelli C. M. (1998) NF-kappaB mediates alphavbeta3 integrin-induced endothelial cell survival. J Cell Biol 141, 1083-93. Schafer M. and Werner S. (2008) Cancer as an overhealing wound: an old hypothesis revisited. Nat Rev Mol Cell Biol 9, 628-38. Schall T. J., Bacon K., Toy K. J. and Goeddel D. V. (1990) Selective attraction of monocytes and T lymphocytes of the memory phenotype by cytokine RANTES. Nature 347, 669-671. Schurch W., Seemayer T. A. and Lagace R. (1981) Stromal myofibroblasts in primary invasive and metastatic carcinomas. A combined immunological, light and electron microscopic study. Virchows Arch A Pathol Anat Histol 391, 125-39.

66 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation Semba S., Kodama Y., Ohnuma K., Mizuuchi E., Masuda R., Yashiro M., Hirakawa K. and Yokozaki H. (2009) Di- rect cancer-stromal interaction increases fibroblast proliferation and enhances invasive properties of scirrhous-type gastric carcinoma cells. Br J Cancer 101, 1365-73. Sen R. and Baltimore D. (1986) Inducibility of kappa immunoglobulin enhancer-binding protein Nf-kappa B by a posttranslational mechanism. Cell 47, 921-8. Senger D. R., Galli S. J., Dvorak A. M., Perruzzi C. A., Harvey V. S. and Dvorak H. F. (1983) Tumor cells secrete a vascular permeability factor that promotes accumulation of ascites fluid.Science 219, 983-5. Serbina N. V., Jia T., Hohl T. M. and Pamer E. G. (2008) Monocyte-mediated defense against microbial pathogens. Annu Rev Immunol 26, 421-52. Serhan C. N., Chiang N. and Van Dyke T. E. (2008) Resolving inflammation: dual anti-inflammatory and pro-resolution lipid mediators. Nat Rev Immunol 8, 349-61. Serini G., Bochaton-Piallat M. L., Ropraz P., Geinoz A., Borsi L., Zardi L. and Gabbiani G. (1998) The fibronectin domain ED-A is crucial for myofibroblastic phenotype induction by transforming growth factor-beta1. J Cell Biol 142, 873-81. Sharma M. (2010) Chemokines and their receptors: orchestrating a fine balance between health and disease.Crit Rev Biotechnol 30, 1-22. Shimomura T., Kondo J., Ochiai M., Naka D., Miyazawa K., Morimoto Y. and Kitamura N. (1993) Activation of the zymogen of hepatocyte growth factor activator by thrombin. J Biol Chem 268, 22927-32. Shimomura T., Miyazawa K., Komiyama Y., Hiraoka H., Naka D., Morimoto Y. and Kitamura N. (1995) Activation of hepatocyte growth factor by two homologous proteases, blood-coagulation factor XIIa and hepatocyte growth factor activator. Eur J Biochem 229, 257-61. Shinomiya N., Gao C. F., Xie Q., Gustafson M., Waters D. J., Zhang Y. W. and Vande Woude G. F. (2004) RNA in- terference reveals that ligand-independent met activity is required for tumor cell signaling and survival. Cancer Res 64, 7962-70. Silzle T., Kreutz M., Dobler M. A., Brockhoff G., Knuechel R. and Kunz-Schurgart L. A. (2003) Tumor-associated fibroblasts recruit blood monocytes into tumor tissue. Eur J Immunol 33, 1311-1320. Siren V., Salmenperä P., Kankuri E., Bizik J., Sorsa T., Tervahartiala T. and Vaheri A. (2006) Cell-cell contact activation of fibroblasts increases the expression of matrix metalloproteinases. Ann Med 38, 212-220. Smith R. S., Smith T. J., Blieden T. M. and Phipps R. P. (1997) Fibroblasts as sentinel cells. Synthesis of chemokines and regulation of inflammation. Am J Pathol 151, 317-22. Solinas G., Germano G., Mantovani A. and Allavena P. (2009) Tumor-associated macrophages (TAM) as major players of the cancer-related inflammation. J Leukoc Biol 86, 1065-73. Sonnenberg E., Meyer D., Weidner K. M. and Birchmeier C. (1993) Scatter factor/hepatocyte growth factor and its receptor, the c-met tyrosine kinase, can mediate a signal exchange between mesenchyme and epithelia during mouse development. J Cell Biol 123, 223-35. Sottile J. (2004) Regulation of angiogenesis by extracellular matrix. Biochim Biophys Acta 1654, 13-22. Strieter R. M. (2002) Interleukin-8: a very important chemokine of the human airway epithelium. Am J Physiol Lung Cell Mol Physiol 283, L688-9. Strieter R. M., Polverini P. J., Kunkel S. L., Arenberg D. A., Burdick M. D., Kasper J., Dzuiba J., Van Damme J., Walz A., Marriott D., Chan S. Y., Roczniak S. and Shanafelt A. B. (1995) The functional role of the ELR motif in CXC chemokine-mediated angiogenesis. J Biol Chem 270, 27348-27357. Tabruyn S. P. and Griffioen A. W. (2008) NF-kappa B: a new player in angiostatic therapy. Angiogenesis 11, 101-6. Tak P. P. and Firestein G. S. (2001) NF-kappaB: a key role in inflammatory diseases. J Clin Invest 107, 7-11. Tanaka Y., Adams D. H. and Shaw S. (1993) Proteoglycans on endothelial cells present adhesion-inducing cytokines to leukocytes. Immunol Today 14, 111-5. Thorpe R., Bacon K., Baggiolini M., Broxmeyer H., Horuk R., Lindley I., Mantovani A., Matsushima K., Murphy P., Nomiyama H., Oppenheim J., Rot A., Schall T., Tsang M., Van Damme J., Wadhwa M., Yoshie O., Zlotnik A. and Zoon K. (2003) Chemokine/chemokine receptor nomenclature. Cytokine 21, 48-49. Tlsty T. D. and Coussens L. M. (2006) Tumor stroma and regulation of cancer development. Annu Rev Pathol 1, 119-50. Tomasek J. J., Gabbiani G., Hinz B., Chaponnier C. and Brown R. A. (2002) Myofibroblasts and mechano-regulation of connective tissue remodelling. Nat Rev Mol Cell Biol 3, 349-63. Tuder R. M., Flook B. E. and Voelkel N. F. (1995) Increased gene expression for VEGF and the VEGF receptors KDR/ Flk and Flt in lungs exposed to acute or to chronic hypoxia. Modulation of gene expression by nitric oxide. J Clin Invest 95, 1798-807. Tzima E., Irani-Tehrani M., Kiosses W. B., Dejana E., Schultz D. A., Engelhardt B., Cao G., DeLisser H. and Schwartz M. A. (2005) A mechanosensory complex that mediates the endothelial cell response to fluid shear stress. Nature 437, 426-31.

References | 67 Umeda K., Ikenouchi J., Katahira-Tayama S., Furuse K., Sasaki H., Nakayama M., Matsui T., Tsukita S. and Furuse M. (2006) ZO-1 and ZO-2 independently determine where claudins are polymerized in tight-junction strand formation. Cell 126, 741-54. Wallez Y. and Huber P. (2008) Endothelial adherens and tight junctions in vascular homeostasis, inflammation and angiogenesis. Biochim Biophys Acta 1778, 794-809. Van Lint P. and Libert C. (2007) Chemokine and cytokine processing by matrix metalloproteinases and its effect on leukocyte migration and inflammation. J Leukoc Biol 82, 1375-81. Wang H. S., Cao H. J., Winn V. D., Rezanka L. J., Frobert Y., Evans C. H., Sciaky D., Young D. A. and Smith T. J. (1996) Leukoregulin induction of prostaglandin-endoperoxide H synthase-2 in human orbital fibroblasts. Anin vitro model for connective tissue inflammation. J Biol Chem 271, 22718-28. Wang H. U., Chen Z. F. and Anderson D. J. (1998) Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 93, 741-53. Wang W., Bergh A. and Damber J. E. (2005) Cyclooxygenase-2 expression correlates with local chronic inflammation and tumor neovascularization in human prostate cancer. Clin Cancer Res 11, 3250-3256. Weidner K. M., Arakaki N., Hartmann G., Vandekerckhove J., Weingart S., Rieder H., Fonatsch C., Tsubouchi H., Hishida T., Daikuhara Y. and Birchmeier W. (1991) Evidence for the identity of human scatter factor and human hepatocyte growth factor. Proc Natl Acad Sci U S A 88, 7001-5. Vestweber D. (2008) VE-cadherin: the major endothelial adhesion molecule controlling cellular junctions and blood vessel formation. Arterioscler Thromb Vasc Biol 28, 223-32. Vestweber D. and Blanks J. E. (1999) Mechanisms that regulate the function of the selectins and their ligands. Physiol Rev 79, 181-213. Wijelath E. S., Murray J., Rahman S., Patel Y., Ishida A., Strand K., Aziz S., Cardona C., Hammond W. P., Savidge G. F., Rafii S. and Sobel M. (2002) Novel vascular endothelial growth factor binding domains of fibronectin enhance vascular endothelial growth factor biological activity. Circ Res 91, 25-31. Wilgus T. A. (2008) Immune cells in the healing skin wound: influential players at each stage of repair. Pharmacol Res 58, 112-6.§ Wojta J., Kaun C., Breuss J. M., Koshelnick Y., Beckmann R., Hattey E., Mildner M., Weninger W., Nakamura T., Tschachler E. and Binder B. R. (1999) Hepatocyte growth factor increases expression of vascular endothelial growth factor and plasminogen activator inhibitor-1 in human keratinocytes and the vascular endothelial growth factor receptor flk-1 in human endothelial cells. Lab Invest 79, 427-38. Xia Y., Pauza M. E., Feng L. and Lo D. (1997) RelB regulation of chemokine expression modulates local inflammation. Am J Pathol 151, 375-87. Xu W. and Kimelman D. (2007) Mechanistic insights from structural studies of beta-catenin and its binding part- ners. J Cell Sci 120, 3337-44. Yannas I. V., Kwan M. D. and Longaker M. T. (2007) Early fetal healing as a model for adult organ regeneration. Tissue Eng 13, 1789-98. Zeine R., Salwen H. R., Peddinti R., Tian Y., Guerrero L., Yang Q., Chlenski A. and Cohn S. L. (2009) Presence of cancer-associated fibroblasts inversely correlates with Schwannian stroma in neuroblastoma tumors. Mod Pathol 22, 950-8. Zen K., Karsan A., Stempien-Otero A., Yee E., Tupper J., Li X., Eunson T., Kay M. A., Wilson C. B., Winn R. K. and Har- lan J. M. (1999) NF-kappaB activation is required for human endothelial survival during exposure to tumor necrosis factor-alpha but not to interleukin-1beta or lipopolysaccharide. J Biol Chem 274, 28808-15. Zhang Y., Cao H. J., Graf B., Meekins H., Smith T. J. and Phipps R. P. (1998) CD40 engagement up-regulates cyclooxygenase-2 expression and prostaglandin E2 production in human lung fibroblasts.J Immunol 160, 1053-7. Zimmerman G. A., Albertine K. H., Carveth H. J., Gill E. A., Grissom C. K., Hoidal J. R., Imaizumi T., Maloney C. G., McIntyre T. M., Michael J. R., Orme J. F., Prescott S. M. and Topham M. S. (1999) Endothelial activation in ARDS. Chest 116, 18S-24S. Östman A. and Augsten M. (2009) Cancer-associated fibroblasts and tumor growth--bystanders turning into key players. Curr Opin Genet Dev 19, 67-73.

68 | Bustle behind the scenes of action: nemosis as a model for fibroblast inflammatory activation