Matti Laaksonen

Associations among Emdogain®, human oral carcinoma cells and

oral proteolytic enzymes

Academic Dissertation

To be presented with the permission of the Faculty of Medicine of the University of Helsinki for public examination in the Lecture hall 3 of Biomedicum Helsinki, Haartmaninkatu 8, Helsinki, on 14th May 2010, at

13 pm.

Department of Cell Biology of Oral Diseases, Biomedicum Helsinki, Institute of Dentistry, Faculty of Medicine,

University of Helsinki

Helsinki 2009

2

Supervised by Professor Tuula Salo, DDS, PhD Department of Diagnostics and Oral Medicine Oulu University Central Hospital Institute of Dentistry University of Oulu Oulu, Finland Professor Timo Sorsa, DDS, PhD, Dipl Perio Department of Cell Biology of Oral Diseases, Biomedicum Helsinki Institute of Dentistry University of Helsinki Helsinki, Finland Rewieved by: Associate Professor Lari Häkkinen, DDS, PhD Department of Oral Biological and Medical Sciences Faculty of Dentistry University of British Columbia Vancouver, BC, Canada Marilena Vered, D.M.D Department of Oral Pathology and Oral Medicine School of Dental Medicine Tel Aviv University Tel Aviv, Israel Opponent: Professor Timo Närhi, DDS, PhD Department of Prosthetic Dentistry and Biomaterials Science Institute of Dentistry University of Turku Turku, Finland ISBN 978-952-10-6198-1 (paperback) ISBN 978-952-10-6199-8 (PDF) Yliopistopaino Helsinki 2010

3

CONTENTS: LIST OF ORIGINAL PUBLICATIONS 4 ABBREVIATIONS 5 ABSTRACT 6 1 INTRODUCTION 7 2 REVIEW OF THE LITERATURE 9

2.1 Healthy and diseased periodontium 9 2.2 Gingival crevicular fluid 10 2.3 Matrix metalloproteinases 11 2.4 Carcinogenesis 14 2.5 Emdogain 18 2.6 Effects of Emdogain on mesenchymal cells 19 2.7 Effects of Emdogain on epithelial cells 22 2.8 Effects of Emdogain on neoplastic cells 22

3 A CASE REPORT 24 4 HYPOTHESES AND AIMS OF THE STUDY 28 5 MATERIALS AND METHODS 29

5.1 Patients and sample collection (I) 29 5.2 Study reagents (I,II,III) 29 5.3 EMD degradation assays (I) 30 5.4 Cell lines and cell cultures (I,II,III) 31 5.5 In vitro cell proliferation, migration, wound closure and adhesion assays (I,II) 31 5.6 In vitro MMP production assay (II, III, unpublished data) 33 5.7 Affymetrix cDNA microarrays (III) 34 5.8 Mouse experiments (II) 35 5.9 Data collection for systematic review (IV) 35 5.10 Statistical analysis (I,II,III) 36

6 RESULTS 38

6.1 Degradation of EMD by GCF and MMP (I) 38 6.2 In vitro cell proliferation, viability and behaviour assays (I, II) 38 6.3 Effects of EMD on the in vitro production of matrix metalloproteinases (II, III, unpublished data) 40 6.4 Effects of EMD and TGF-β1 on the gene profile of carcinoma cells (III) 41 6.5 Effects of EMD on metastasis formation in vivo (II) 42

6.6 Results of the systematic review (IV) 42 7 DISCUSSION 44

7.1 Summary of the main results 44 7.2 EMD in periodontal regeneration 44

7.3 EMD and carcinogenesis 46 7.4 Methodological considerations 50 8 CONCLUSIONS 54 ACKNOWLEDGEMENTS 55 REFERENCES 57 APPENDIX 70

4

LIST OF ORIGINAL PUBLICATIONS

I Laaksonen M, Salo T, Vardar-Sengul S, Atilla G, Han Saygan B, Simmer JP, Baylas H, Sorsa

T. Gingival crevicular fluid can degrade Emdogain and inhibit Emdogain induced

proliferation of periodontal ligament fibroblasts. Journal of Periodontal Research 2009, Nov 9

Epub ahead of print (DOI: 10.1111/j.1600-0765.2009.01244.x.)

II Laaksonen M, Suojanen J, Nurmenniemi S, Läärä E, Sorsa T, Salo T. The enamel matrix

derivative (Emdogain) enhances human tongue carcinoma cells gelatinase production,

migration and metastasis formation. Oral Oncology 2008;44:733-742

III Laaksonen M, Sorsa T, Vilen ST, Salo T. Emdogain and TGF-β1 target genes in human

tongue carcinoma cells in vitro. Submitted

IV Laaksonen M, Sorsa T, Salo T. Emdogain in carcinogenesis: a systematic review of in vitro

studies. Journal of Oral Science 2010;52:1-11

5

ABBREVIATIONS

ALP alkaline phosphatase BM basement membrane BMP bone morphogenetic protein BMSC bone marrow stromal cells BrdU 5-bromo-2’-deoxyuridine BSA bovine serum albumin BSP bone sialoprotein cAMP cyclic adenosine monophosphate CI confidence interval COX cyclo oxygenase CTGF connective tissue growth factor CTT-2 gelatinase inhibitor, (GRENYHGCTTHWGFTLC) DMEM Dulbeccos Modified Eagles’s Medium DNA deoxyribonucleic acid ECL enhanced chemiluminescence ECM extra cellular matrix EDTA ethylendiamine tetraacetic acid EGF epidermal growth factor ELISA enzyme linked immunosorbent assay EMD Emdogain® EMP enamel matrix protein EMT epithelial to mesenchymal transition ERK extracellular signal-regulated kinase ERM epithelial cell rests of Malassez FC fold change FGF fibroblast growth factor GCF gingival crevicular fluid GO gene ontology HNSCC head and neck squamous cell carcinoma IGF insulin-like growth factor IL interleukin MAP mitogen-activated protein MMP matrix metalloproteinase OPG osteoprotegerin PBS phosphate buffered saline PDGF platelet-derived growth factor PGE2 prostaglandin E2 PLF periodontal ligament fibroblast RANKL receptor activator of NF-kB ligand RNA ribonucleic-acid RT room temperature RTK receptor tyrosine kinase SCC squamous cell carcinoma SD standard deviation SDS-PAGE sodium dodecyl sulphate polyacrylamide gel electrophoresis TCA trichloroacetic acid TGF-β transforming growth factor-β TIMP tissue inhibitor of matrix metalloproteinase TNC Tris-NaCl-CaCl2-buffer TNF-α tumour necrosis factor-α VEGF vascular endothelial growth factor

6

ABSTRACT

The Enamel matrix derivative Emdogain® (EMD) is a commercially available tissue extract preparation of porcine enamel origin. Studies have shown EMD to be clinically useful in promoting periodontal regeneration. EMD has been widely used in periodontal therapy for over ten years, but the mechanism of its action and the exact composition are not completely clear. EMD is predominantly amelogenin (>90%). However, unlike amelogenin, EMD has a number of growth factor-like effects and it has been shown to enhance the proliferation, migration and other cellular functions of periodontal ligament fibroblasts and osteoblasts. In contrast, the effects of EMD on epithelial cell lines and in particular on oral malignant cells have not been adequately studied. In addition, EMD has effects on the production of cytokines by several oral cell lines and the product is in constant interaction with different oral enzymes. Regardless of the various unknown properties of EMD, it is said to be clinically safe in regenerative procedures, also in medically compromised patients. The aim of the study was to examine whether gingival crevicular fluid (GCF), which contains several different proteolysis enzymes, could degrade EMD and alter its biological functions. In addition, the objective was to study the effects of EMD on carcinogenesis-related factors, in particular MMPs, using in vitro and in vivo models. This study also aimed to contribute to the understanding of the composition of EMD. GCF was capable of degrading EMD, depending on the periodontal status, with markedly more degradation in all states of periodontal disease compared to healthy controls. EMD was observed to stimulate the migration of periodontal ligament fibroblasts (PLF), whereas EMD together with GCF could not stimulate this proliferation. In addition, recombinant amelogenin, the main component of EMD, decreased the migration of PLFs. A comparison of changes induced by EMD and TGF-β1 in the gene profiles of carcinoma cells showed TGF-β1 to regulate a greater number of genes than EMD. However, both of the study reagents enhanced the expression of MMP-10 and MMP-9. Furthermore, EMD was found to induce several factors closely related to carcinogenesis on gene, protein, cell and in vivo levels. EMD enhanced the production of MMP-2, MMP-9 and MMP-10 proteins by cultured carcinoma cells. In addition, EMD stimulated the migration and in vitro wound closure of carcinoma cells. EMD was also capable of promoting metastasis formation in mice. In conclusion, the diseased GCF, containing various proteases, causes degradation of EMD and decreased proliferation of PLFs. Thus, this in vitro study suggests that the regenerative effect of EMD may decrease due to proteases present in periodontal tissues during the inflammation and healing of the tissues in vivo. Furthermore, EMD was observed to enhance several carcinoma-related factors and in particular the production of MMPs by benign and malignant cell lines. These findings suggest that the clinical safety of EMD with regard to dysplastic mucosal lesions should be further investigated.

7

1 INTRODUCTION

Repair and regeneration of periodontal tissues are major goals in periodontal treatment. Periodontal

regeneration is a complex multifactorial process in which molecular interactions in and between

mesenchymal and epithelial cell lines are needed (Nyman et al 1982, Pitaru et al 1994). The

essential stages in successful regeneration are cementogenesis, osteogenesis and assembly of

periodontal ligaments onto the cementum. At the same time, the apical proliferation of epithelial

cells must be arrested (Gottlow et al 1984, Ripamonti & Reddi 1997, Zeichner-David 2006). In

1997, extracted porcine enamel matrix proteins were observed to be capable of inducing new

cementum and bone formation in periodontal defects in monkeys (Hammarström 1997). Since then,

this tissue extract preparation, consisting of hydrophobic enamel matrix proteins from developing

porcine enamel, has been marketed under the commercial name of Emdogain® (EMD). Several in

vivo studies since then have shown EMD to be clinically useful in promoting periodontal

regeneration, including the restoration of alveolar bone, cementum and collagenous ligaments

(Esposito et al 2004).

Although EMD is at present widely used in periodontal therapy, the mechanisms of action and the

exact composition of the product are not completely clear. The predominant compound of EMD is

amelogenin (>90%) (Zeichner-David 2006). In spite of amelogenin not being a growth factor, EMD

possesses several growth factor-like effects. Therefore, many attempts have been made to identify

growth factors in EMD, without satisfying results. However, there is abundant evidence regarding

the effects of EMD on periodontal tissues. EMD can stimulate the migration and proliferation of

periodontal fibroblasts and osteoblasts and induce the production of several cytokines in these cells.

In contrast to periodontal fibroblasts and osteoblasts, other cell lines of the oral cavity, particularly

malignant cell lines, have not been adequately studied. Despite the lack of studies and the fact that

EMD is known to have growth factor-like effects on different cell lines, EMD is said to be

clinically safe (Zetterström et al 1997, Heard et al 2000, Oringer 2002). In studies, however, the

clinical safety has often entailed only the capability of porcine enamel proteins to induce

immunological reactions or local post-operative symptoms such as pain or sensitivity in the teeth

involved. Nonetheless, EMD is applied in high doses directly in periodontal pockets where it

interacts with proteolytic enzymes, e.g. MMPs, of inflamed tissue. Thus, this combination of

proteolytic enzymes together with unknown cytokines of EMD might also be capable of inducing

unwanted effects and alterations in vivo.

8

These observations led to two distinct hypothesis of the study. The first hypothesis was that the

proteases present in oral tissues and oral fluids can degrade EMD and the degradation alters the

effects of EMD in vitro. The Second hypothesis was that EMD could be capable of inducing

carcinogenesis. Thus, the purpose of this study was to examine whether gingival crevicular fluid,

containing several different proteolytic enzymes, could degrade EMD and alter the proliferation of

periodontal fibroblasts. In addition, the study aimed to determine if EMD could induce

carcinogenesis-related factors, in particular MMPs, in vitro and in vivo. Furthermore, this study

aimed to contribute to the understanding of the composition of EMD.

9

2 REVIEW OF THE LITERATURE

2.1 Healthy and diseased periodontium

Periodontal tissue surrounding the tooth consists of four different developmental, biological and

functional units: gingiva, root cementum, alveolar bone and periodontal ligament (Mueller 2005).

These tissues can support and nourish the related tooth and protect underlying tissues, and thus are

responsible for the main functions of the periodontium.

Figure 1. The structure of the periodontium. A. corresponds to a healthy state and B. to periodontitis. The vertical arrows show the flow direction of the gingival crevicular fluid to the gingival crevice.

The transition from a healthy state or gingivitis to periodontitis is characterised by proteolytic

destruction of periodontal ligament (type I collagen fibers) and alveolar bone accompanied by

apical migration of sulcular epithelial cells. This phenomenon leads to periodontal pocket formation

and eventually to loss of the tooth involved. Periodontal diseases can be grouped according to the

degree, severity and activity of the tissue destruction. In 1999, an International Workshop for the

Classification of Periodontal Disease and Conditions was organised to revise the classification

system of periodontal diseases (Armitage 1999). The major disease categories are: I gingival

10

disease, II chronic periodontitis, III aggressive periodontitis, IV periodontitis as a manifestation of

systemic diseases, V necrotising periodontal disease, VI abscesses of the periodontium, VII

periodontitis associated with endodontic lesions, and VIII development of acquired deformities and

conditions.

All the forms of periodontitis are first and foremost infectious diseases and initiated by bacteria. All

of the periodontopathogens, particularly Porphyromonas gingivalis, Aggregatibacter

actinomycetemcomitans and Treponema denticola, can produce proteolytic enzymes directly

capable of degrading host tissues and activating host pro-MMPs (Ding et al 1995, Potempa et al

2000, Eley & Cox 2003). However, host-derived proteolytic enzymes cause most of the tissue

destruction in periodontitis. Lipolysaccarides produced by gram-negative bacteria stimulate

polymorphonuclear leukocytes and macrophages to secrete inflammatory products such as

prostaglandin E2 (PGE2), interleukin-1 (IL-1), -6 and -8, tumour necrosis factor alpha (TNFα), and

matrix metalloproteinases (MMPs). These products in turn activate vascular smooth muscle cells,

sulcular epithelial cells, gingival fibroblasts, polymorphonuclear leukocytes, macrophages, mast

cells and osteoclasts to stimulate the inflammation process further by producing various proteolytic

enzymes capable of degrading periodontal soft-tissues and alveolar bone (Birkedal-Hansen 1993,

Sorsa et al 2004, 2006).

2.2 Gingival crevicular fluid

Healthy gingival crevicular fluid (GCF) is a transudate of serum origin that diffuses passively

through the gingival tissues, BM, and epithelium to the gingival crevice (Cimosoni 1983, Curtis et

al 1990). In a healthy gingiva, GCF is secreted only in small amounts and is rather homogenous,

containing mainly plasma components and normal tissue turnover-related proteins (Curtis et al

1990). Due to plague formation and bacterial accumulation around the tooth, the vascular

permeability of gingival capillaries increases according to the level of the disease, i.e. healthy

gingiva, gingivitis and periodontitis. As a result, the GCF changes from transudate into an

inflammatory exudate and the volume secreted increases. Furthermore, the protein content of GCF

in diseased states differs significantly from that of the healthy GCF. Diseased GCF contains more

proteolytic enzymes as well as proteins of non-plasma origin produced by the cells of the

periodontal connective tissue and by the sulcular and attached epithelium of the gingiva (Curtis et al

1990, Sorsa et al 2004, 2006). There are more than 65 different infection-induced enzymes found in

GCF, including gelatinases, collagenases, cytokines and various tissue breakdown products

11

(Armitage 2004). In addition to different proteins, diseased GCF is also rich in cells, in particular

leukocytes and microbes from the bacterial plaque (Lamster & Ahlo 2007). Regardless of the

increased flow rate of the GCF in periodontitis, the volume alone does not necessarily reflect

periodontal disease activity. However, the enzyme composition of GCF follows and reflects the

progression of periodontitis and may serve as a diagnostic tool of the disease (Kinane et al 2003,

Sorsa et al 2004, 2006, Mäntylä et al 2003, 2006).

2.3 Matrix metalloproteinases

Structural and functional characteristics of matrix metalloproteinases

Human endopeptidases can collectively degrade all extracellular matrix (ECM) and basement

membrane (BM) components and are pivotal to normal tissue turnover and development as well as

to several destructive diseases, including periodontitis and carcinoma. They can be divided into five

major classes, i.e. serine, cysteine, metallo, aspartic and thereonine proteases, based on their

catalytic properties and inhibitor sensitivities (Stöcker & Bode 1995, Cawston & Wilson 2006,

Lόpez-Otίn & Matrisian 2007). Matrix metalloproteinases were discovered in 1962 to be

responsible for the observed degradation of the frog tail. Since then, 23 human matrix

metalloproteinases (MMPs) have been found and they form a superfamily of structurally related

zinc-dependent endopeptidases. MMPs are either secreted or cell-membrane bound, and they are

most often classified according to their structure and substrate specificity into six subgroups:

collagenases (MMP-1, -8, and -13), gelatinases (MMP-2 and -9), stromelysins and stromelysin-like

MMPs (MMP-3, -10, -11, and -12), matrilysins (MMP-7 and -26), membrane-type MMPs (MMP-

14, -15, -16, -17, - 24, and -25), and other MMPs (MMP-19, -20, -21, -23, -27, and -28) (Nagase &

Woessner 1999, Sternlicht & Werb 2001, Nagase et al 2006). The MMP polypeptides differ in size

(20 to 100 kDa) and contain several distinct domains, which are similar among the different

members of the family. The basic domain structure of MMPs consist of a pre-domain, a zinc-

binding catalytic domain, and a C-terminal hemopexin domain (excluding MMP- 7 and -26), which

is linked to the catalytic domain by a hinge region. They are, with the exception of MMP-11,

secreted as inactive zymogens and activated outside the cell by other MMPs, serine proteases,

microbial proteases as well as other factors such as oxygen-derived free radicals (Saari et al 1990,

Sorsa et al 1992, Moilanen et al 2003). For their activation, proteolytic removal of the propeptide-

domain is required. This results in a reduction in the molecular weight and enables access to the

catalytic site of the MMPs (Nagase 1997, Nagase et al 2006, Page-McCaw et al 2007).

12

MMPs act as the main processors of ECM components. As such, they participate in physiological

processes involving ontogenesis, organ development, tissue turnover and repair. Thus, MMPs are

centrally involved in e.g. ovulation, embryogenesis, organ morphogenesis, nerve growth,

angiogenesis, hair follicle cycling, inflammatory cell function, apoptosis, tooth eruption, bone

remodelling and wound healing. More recently, MMPs have also been implicated in more

sophisticated processes than mere ECM turnover. These include the release, activation and

modification of cell-signalling molecules and their receptors. Substrates in these processes can vary,

and can be other proteases, protease inhibitors, clotting factors, antimicrobial peptides, growth

factors, hormones, and cytokines (Brinckerhoff & Matrisian 2002, Egeblad & Werb 2002, Mott &

Werb 2004, Murphy & Nagase 2008).

Under physiological conditions there is a precise regulation of proteolytic degradation and

inhibition of proteolysis. To ensure physiologically beneficial MMP activity, regulation takes place

at all levels of protein synthesis (e.g. transcription, translation) and secretion. In addition, MMP

activity is regulated by physiological inhibitors, i.e.TIMPs (Visse & Nagase 2003). Furthermore,

regulatory actions are mediated by growth factors, hormones, cytokines, cell matrix, cell-cell

interactions and cellular transformation as well as by extrinsic chemical/physical factors.

Disturbance in this physiological balance seems to be related to several pathological conditions and

destructive diseases, including cancer and periodontal disease, together with arthritis,

cardiovascular disease, nephritis, neurological diseases, lung diseases, gastric ulceration, eye

diseases, corneal ulceration, skin diseases, liver fibrosis, atherogenesis, emphysema and chronic

wound healing (Nagase & Woessner 1999, McCawley & Matrisian 2000, Sorsa et al 2004, Nagase

et al 2006).

Gelatinases; MMP-2 and MMP-9

Gelatinases (type IV collagenases) MMP-2 and MMP-9 are also called gelatinase A and gelatinase

B, respectively, because they were first identified based on their ability to degrade gelatin. The

structure of the two gelatinases is similar and typical for the MMPs. Both MMP-2 and MMP-9 also

contain fibronectin type II-like repeats in their catalytic domains. Due to these repeats, the enzymes

have largely overlapping substrate specificities and high affinities for type IV collagen, laminins

and elastin (Van den Steen 2002, Björklund & Koivunen 2005). In addition, gelatinases can cleave

a number of other ECM components as well, including type II, III, V, VII and X collagens,

13

fibronectin, fibrillin and aggrecan (Birkedal-Hansen 1993, Ravanti & Kähäri 2000). However, only

MMP-2 can degrade type I collagen (Björklund and Koivunen 2005). MMP-2 is secreted as a 72-

kDa pro-form that is cleaved into a 59-64-kDa active form (Salo et al 1983). The corresponding

pro- and active forms of MMP-9 have masses of 92 kDa and 68-83 kDa, respectively (Birkedal-

Hansen 1993).

Despite their similarities in structure and substrate profile, MMP-2 and MMP-9 are regulated quite

differently from each other at the level of gene transcription. In general, MMP-2 is produced

constitutively and by a wide range of cell types, i.e. dermal fibroblasts, keratinocytes, endothelial

cells and macrophages (Frisch & Morisaki 1990, Birkedal-Hansen 1993). Expression of MMP-9 is

normally more restricted, but inflammatory stimulation can lead to increased MMP-9 expression in

many cell types including neutrophils, endothelial cells, alveolar cells, macrophages, fibroblasts,

and other connective tissue cells (Vartio et al 1982, Naguse & Woessner 1999, van den Steen et al

2002).

At present, gelatinases are intensively studied, and due to their ability to degrade major BM

components they have been found to be crucial for carcinogenesis including tongue squamous cell

carcinomas (SCCs). Activities of these MMPs have also been found to correlate with the invasive

potential and bad prognosis of cancer in e.g. head and neck SCCs (Salo et al 1982, Sutinen et al

1998, Thomas et al 1999, Hong et al 2000, Ruokolainen et al 2004, de Vicente et al 2005). In

periodontitis, particularly MMP-9 seems to be related to tissue destruction and progression of the

disease as well as being one of the predominant MMPs present in inflamed gingival tissue, gingival

crevicular fluid, saliva-samples and dental peri-implant sulcular fluid (Sorsa et al 2004, 2006).

Stromelysins; MMP-10

The stromelysin subfamily consists of stromelysin-1 (MMP-3) and stromelysin-2 (MMP-10).

Stromelysin-3 (MMP-11) and human macrophage metalloelastase (MMP-12) are often included in

this subgroup as being stromelysin-like MMPs (McCawley & Matrisian 2001, Visse and Nagase

2003, Sorsa et al 2006). MMP-3 and MMP-10 are expressed in vivo by keratinocytes and fibroblasts

as well as several malignant cell types (Birkedal-Hansen et al 1993, Giambernardi et al 1998,

Saarialho-Kere 1998, Rechardt et al 2000). Their main substrates are basement membrane

components such as type IV, V, IX and X collagen, fibronectin, laminin, elastin, gelatin and

proteoglycan core proteins, but not interstitial collagens (Sorsa et al 2006). Stromelysins can also

14

activate other proMMPs. More precisely, MMP-10 activates proMMPs -1, -2, -7, -8, and -9,

whereas MMP-3 activates proMMPs -1, -3, -7, -8, -9, and -13 (Nicholson et al 1989, Windsor et al

1993, Knäuper et al 1996, Nakamura et al 1998, Kähäri & Saarialho-Kere 1999, McCawley &

Matrisian 2001). Stromelysins themselves are activated by several factors including plasmin,

kallikrein, chymase, elastase, and cathepsin G (Nagase & Woessner 1999; Lijnen 2001).

Beside their physiological functions, stromelysins, like other MMPs, are related to pathological

conditions. MMP-10 over-expression in transgenic mice leads to disorganised cell migration during

wound healing (Krampert et al 2004). Hence, over-expression may lead to impaired wound healing,

whereas normal regulation promotes epithelial repair. In addition, MMP-10 is also widely

expressed in several tumours of epithelial origin, including oral carcinomas (Birkedal-Hansen et al

2000, Impola et al 2004), where the tumour cells themselves express MMP-10, unlike several other

MMPs that are localised predominantly in tumour stroma (Gill et al 2004).

2.4 Carcinogenesis

Background on oral carcinomas

Head and neck squamous cell carcinomas (HNSCCs), arising from the mucosal lining of the oral

cavity and pharynx, are currently the sixth most prevalent neoplasm in the world, with

approximately 900 000 cases diagnosed worldwide annually (Chin et al 2006). In Finland, 481 new

cases of HNSCCs were diagnosed in 2007. Of these, 120 were squamous cell carcinomas of the

tongue and the prevalence of 56 new cases in women was the highest since 1961 (Finnish Cancer

Registry). Due to the late stage of diagnosis, and to a high frequency of metastasis, recurrence and

second primary tumours, the survival has remained poor and unchanged during the last three

decades. In addition, HNSCC is also one the most distressing of all cancers, since extensive surgery

is needed at the site of the most complex functional anatomy in the human body. Thus, the therapies

often cause serious paralyses in the patient along with a need for extensive regenerative and

rehabilitation procedures.

Mechanisms of cancer growth, invasion and metastasis

Tumour development and progression is a complex multistep process in which genetic errors are

required for the first malignant cell to evolve, normally through several pre-malignant alterations.

15

This malignant cell needs to be hidden from the immune system and lose its normal proliferative

control, and be able to divide rapidly to form tumour tissue and to invade surrounding tissues and

eventually distant organs. Thus, several modifications in the environment that favour malignant

progression are needed (Rusciano & Burger 1992, Fidler 1995, Hanahan & Weinberg 2000).

One of the first steps towards malignancy is the epithelial to mesenchymal transition

(EMT), resulting in free movement of the formerly tightly-bound epithelial cells. The

ECM works as a barrier to free cell movement, and in order to migrate, cells need to activate their

cytoskeletal motors for movement, modulate their cell-surface adhesion molecules for traction,

cleave ECM to break the barrier and respond to chemoattractants for guidance in migration (Vu &

Werb 2000). The next and most crucial step of carcinogenesis is the invasion through underlying

BM, beyond which the tumour cells can spread into the adjacent connective tissue relatively freely

(Hanahan & Weinberg 2000). In this invasion process, the proteolytic degradation of basement

membrane and extracellular structures combined with enhanced cell motility are essential (Friedl &

Wolf 2003). The next goal of the tumour is to form metastasis. First malignant cells penetrate

through the walls of blood vessels or lymphatic vessels. Then, a clot of cells travels with the stream,

attaches to the wall of a distant vessel, moves through BM and ECM and begins to grow again at

the new site.

After the initial formation, the primary tumour can reach a maximum size of only 1-2 mm without a

blood supply. Thus, the continual growth and survival of a solid tumour requires angiogenesis.

Angiogenesis results from hypoxia-induced expression of basic fibroblast growth factor (bFGF) and

vascular endothelial growth factor (VEGF), growth factors that promote endothelial cell recruitment

(Wyke 2000). Interestingly, MMPs are reported to release these factors into the tumour stroma

(Folkman 2007). Tumour neovasculature leaks more than normal vasculature, thus it is more

efficient in the delivery of oxygen to the tumour (Sternlicht et al 1999, Folkman 2006).

Angiogenesis has been linked to increased metastasis formation and decreased survival of patients

with HNSCC (Sauter et al 1999, Smith et al 2000).

We have only recently started to understand the factors that affect the tumour microenvironment

and homeostasis. Cells of SCC develop molecular strategies in order to evade the growth inhibitory

effects of cytokines present in the tumour microenvironment. Therefore, the malignant

transformation process is strongly associated with an altered response to cytokine stimulation

(Douglas et al 2004). A tumour induces or spontaneously produces several immune suppressive

16

mediators causing immune dysfunctions (Chin et al 2004). Human HNSCC cells have been shown

to secrete high levels of numerous cytokines involved in indirect modulation mechanisms of

immune responses and proangiogenic processes (Pries et al 2006). Interestingly, the stroma

surrounding the tumour can also participate in carcinogenesis. Stromal fibroblasts and inflammatory

cells can cause tumorigenic conversion of epithelial cells. In addition, stromal cells can participate

in the proteolysis of ECM and BM by producing proteolytic enzymes, including MMPs.

Furthermore, stroma can already support and promote angiogenesis when the BM is still intact

(Hanahan & Folkman 1996, Westermarck & Kähäri 1999, Stetler-Stevenson & Yu 2001, Friedl &

Wolf 2003).

MMPs -2, -9 and -10 in carcinogenesis

A large variety of different proteases have been found to be associated with cancer progression but

MMPs are considered to be particularly crucial in this process (McCawley & Matrisian 2000,

Björklund & Koivunen 2005). Co-expression of several members of the MMP family is

characteristic of human malignant tumours (van Kempen et al 2002). Elevated MMP expression has

been found to be related to tumour invasion properties and patient prognosis. Classical MMPs have

an important role at all stages of tumorigenesis: they enhance tumour-induced angiogenesis, release

growth factors from the matrix or cleave their receptors, regulate apoptosis, and break down ECM

and BM to allow tumour cell invasion and metastatic spread (McCawley & Matrisian 2000, Liotta

& Kohn 2001, Stetler-Stevenson & Yu 2001). Paradoxically, and depending on cancer type, several

MMPs (MMPs-3, -7, - 9, and -12) can also possess anti-cancer effects by acting as anti-

inflammatory or tumour suppressor agents or by releasing angiogenic inhibitors, such as angiostatin

and endostatin (Cornelius et al 1998, Nyberg et al 2003, 2008). During carcinogenesis, MMPs are

produced by malignant cells, but in many cases the greater portion of MMPs are actually secreted

by stromal cells. For example, gelatinases (MMP-2 and MMP-9) are produced in vitro by neoplastic

keratinocytes (Lyons et al 1991, Salo et al 1991). However, it is assumed that in vivo MMP-2 is

produced mainly by stromal fibroblasts, whereas MMP-9 is produced both by stromal and

carcinoma cells (Impola et al 2004, Graesslin et al 2006). Thus, the processes leading to MMP

production, proteolysis and tumour growth are dependent on interactions in and between stromal

and malignant cells, and these actions are probably precisely controlled and highly organised.

The high ability of gelatinases (MMP-2 and MMP-9) to destroy BM (type IV collagen) and ECM

structures may explain why they are often associated with tumour progression and with metastatic

17

formation of several carcinomas and in particular HNSCCs (Hong et al 2000, Franchi et al 2002, de

Vicente et al 2005, Björklund & Koivunen 2005). In addition to BM and ECM degradation and

metastasis, gelatinases take part in neoangiogenesis. MMP-9 seems to be especially associated with

tumour vascularisation, since MMP-9 deficient mice have been shown to fail to grow

neovasculature, whereas this ability was regained after transplantation of MMP-9 producing wild-

type stroma (Ahn & Brown 2008). Furthermore, MMP-2 and MMP-9 have been found to be up-

regulated in angiogenic lesions and to up-regulate the release of VEGF, which is a major promoter

of angiogenesis (Bergers et al 2000, Giraudo 2004, van Hinsbergh & Koolwijk 2008). In contrast to

these observations, MMP-9 has also been found to be able to release angiostatin from the carcinoma

stroma, resulting in diminished neovascularisation (Bendrik et al 2008). Thus, the roles of

gelatinases in neoangiogenesis are dual and at present not completely clear. Gelatinases can also

regulate inflammatory reactions and apoptosis in carcinomas (Wielockx et al 2001). MMP-9 can

suppress the proliferation of T lymphocytes by cleaving the interleukin-2a receptor, and regulate

immune reactions by activating TGF-β and by cleaving intercellular adhesion molecule 1 (ICAM-

1), making tumour cells resistant to natural killer cell-mediated cytotoxicity (Sheu et al 2001, Fiore

et al 2002). On the other hand, MMP-2 is able to proteolytically process monocyte chemoattractant

proteins and suppress inflammation in vivo (McQuibban et al 2002). Clinically, MMP-2 and MMP-

9 are associated with aggressive forms of head and neck cancers with poor prognosis (Sutinen et al

1998, Ruokolainen et al 2004, 2006, de Vicente et al 2005). In a recent meta-analysis, the relative

risk for progression of MMP-9 positive cases from oral dysplasia to cancer was 19.00, whereas the

relative risk factor for MMP-2 was 2.9 and considered insignificant by the authors (Smith et al

2009).

Unlike several other MMPs that are predominantly produced by tumour stroma, MMP-10 is

expressed by malignant cells themselves (Nakamura et al 1998, Kerkelä et al 2001). MMP-10 is

over-expressed in several human tumours of epithelial origin, including gastric cancer (Aung et al

2006), esophageal carcinoma (Mathew et al 2002), lung carcinoma (Gill et al 2004), skin carcinoma

(Kerkelä et al 2001) and bladder transitional cell carcinoma (Seargent et al 2005). In oral

carcinomas, MMP-10 is widely expressed but rarely in adjacent oral tissues (Nakamura et al 1998,

Birkedal-Hansen et al 2000). The expression profile of MMP-10 in several human tumours, coupled

with broad substrate specificity and expression in several stages of the tumorigenic process,

suggests that MMP-10 could be a potential tumour marker and a predictor of clinical behaviour

(Duffy et al 2000, Kerkelä et al 2001, Vihinen et al 2002, Gill et al 2004).

18

TGF-β in carcinogenesis

In humans, TGF-β has been identified in three different isoforms, TGF-β1, -β2 and -β3. Of these,

TGF-β1 is the most abundant in most tissues including epithelium. It is a key regulator of epithelial

homeostasis and also acts as one of the major cytokines in carcinogenesis (Derynck et al 2001,

Prime et al 2004, Rosenthal et al 2004). TGF-β is released by tumour cells in an inactive latent

form, which in turn can be activated by several factors including MMP-2 and MMP-9 (Yu &

Stamenkovic 2000, Jenkins 2008). During tumour development, TGF-β has a dual role. At the early

stage of carcinogenesis or in benign tissues it is a potent tumour suppressor and inhibits cell growth.

However, once tumorigenesis is in later stages, TGF-β stimulates cell growth, angiogenesis,

invasion and metastasis, including in head and neck squamous cell carcinomas (HNSCC) (Derynck

et al 2001, Akhurst et al 2001). Whereas MMPs-2 and -9 are capable of activating TGF-β, TGF-β

can in turn induce expression of several MMPs in carcinoma and endothelial cells, including MMP-

1, -2, -9 and -10 (Salo et al 1994, Johansson et al 2000, Wilkins-Port & Higgins 2007, Sinpitaksakul

et al 2008). Furthermore, TGF-β can induce epithelial to mesenchymal transition in squamous cell

carcinomas, resulting in fibroblastoid or ‘spindle-cell’ tumours having highly malignant and

invasive properties (Derynck et al 2001). The mechanisms of angiogenesis stimulation by TGF-β

presumably combine direct and indirect effects. TGF-β induces expression of VEGF, which in turn

stimulates angiogenesis (Pertovaara et al 1994). TGF-β can also directly induce capillary formation

of endothelial cells cultured on a collagen matrix (Madri et al 1988). Taken together, this

information suggest that TGF-β is one of the major factors in carcinogenesis and correlates with

invasive phenophypes of carcinomas (Cui et al 1996, Hagedorn et al 1999, Pasini et al 2001).

2.5 Emdogain

Composition of Emdogain

Findings showing that enamel matrix proteins (EMPs) are expressed along the forming root

suggested that EMPs may have a pivotal role in the differentiation of progenitor cells into

cementoblasts (Slavkin 1989). In 1997, it was discovered that the extracted porcine enamel matrix

proteins could actually induce new cementum and bone formation in periodontal defects in

monkeys (Hammarström et al 1997). Since then, this tissue extract preparation of porcine origin has

19

been marketed under the brand name of Emdogain® (EMD). Commercial Emdogain consist of

enamel matrix derivate, water, and a carrier, i.e. propylene glycol alginate (Bosshardt 2008). The

exact composition of enamel matrix derivative is not completely clear but it contains, at least in

part, the same proteins as those secreted by Hertwig’s epithelial cells. Thus, amelogenin is the main

protein in EMD, constituting over 90% of the product. In addition, EMD contains non-amelogenin

proteins including enamelin, tufelin and ameloblastin (Zeichner-David 2001, 2006). EMD has been

found to be more effective than amelogenin in enhancing fibroblast proliferation together with other

cellular functions (Hoang et al 2002, Chong et al 2006). Therefore, it is generally assumed that

EMD also contains other biologically active factors beside enamel proteins (Suzuki et al 2005,

Chong et al 2006). Kawase et al (2001, 2002) and later Suzuki et al (2005) detected TGF-β1- or

TGF-β-like substances in EMD, and suggested that they are the main functional components of the

product. In addition, EMD has been documented to contain a BMP-like growth factor, which

belongs to the TGF-β family, and also BSP-like molecules (Iwata et al 2002, Suzuki et al 2005).

However, contradictory results have also been obtained. Therefore, the cytokine content of EMD

has not been confirmed and warrants further study (Gestrelius et al 1997).

Clinically, EMD is widely used for periodontal regeneration of a teeth affected by periodontitis.

Experimentally, EMD has also been used for tooth replantation, dentin repair, tooth movement and

skin wound healing. Despite extensive use and numerous studies, a recent Cochrane review

suggests that the overall treatment effect of EMD might be overestimated and the actual clinical

advantages of the product are unknown (Esposito et al 2005).

2.6 Effects of Emdogain on mesenchymal cells

Emdogain and bone-derived cells

Alveolar bone cells, e.g. osteoblasts, osteoclasts and their precursors, are crucial in periodontal

regeneration. During the development of the periodontium, dental follicle cells can differentiate into

periodontal ligament fibroblasts, cementoblasts and osteoblasts (Slavkin et al 1989, Hammarström

et al 1997, Spahr et al 1999, Hakki et al 2001, Zeichner- David 2001). EMD has been shown to be

capable of stimulating the proliferation of follicle cells (Hakki et al 2001). In osteoblastogenesis,

EMD has been shown to promote osteogenic differentiation of pluripotential mesenchymal cell

lines into the osteoblast and/or chondroblast lineage and to stimulate transcription factors related to

20

osteogenesis and chondrogenesis (Ohyama et al 2002, Narukawa et al 2007). Furthermore, EMD

enhances the proliferation and osteogenic potential of bone marrow stromal cells (BMSCs) (Keila et

al 2004, Guida et al 2007). In addition, EMD has been observed to increase the ability of BMSCs’

to differentiate into cementoblasts (Song et al 2007). In contrast, EMD has been observed to have

no effect on cell growth and differentiation of rat bone marrow cells (Gurpinar et al 2003, van den

Dolder et al 2006).

The studies have concluded that EMD stimulates growth, proliferation and mobility of mature

osteoblastic cell lines and cementoblasts and regulates their gene expression (Tokiyasu et al 2000,

Jiang et al 2001, He et al 2004, Schwarz et al 2004, Shimizu et al 2004, Galli et al 2006, Carinci et

al 2006, Klein et al 2006, Jiang et al 2006). EMD enhances the differentiation of the osteoblasts and

the expression of phenotype markers by increasing the production of BSP, cellular alkaline

phosphatase (ALP) and osteocalcin (Tokiyasu et al 2000, Schwartz et al 2000, Hägewald et al 2004,

Shimizu et al 2004, Galli et al 2006). These observations suggest that EMD stimulates proliferation

at an early stage of osteoblastic maturation and enhances differentiation in later stages of cell

development.

EMD also has effects on cytokines produced by osteoblasts. EMD has been observed to stimulate

TGF-β1, connective tissue growth factor (CTGF) and fibroblast growth factor (FGF-2) expression

in human osteoblastic cells (Schwartz et al 2000, Mizutani et al 2003, Heng et al 2007). In addition,

increased COX-2 and decreased MMP-1 expression have been demonstrated (Mizutani et al 2003).

Both EMD and TGF-β1 can protect osteoblasts from inflammation-induced apoptosis (He et al

2005). In the endochondral pathway, EMD can stimulate chondrocyte proliferation in addition to

chondrogenic differentiation (Dean et al 2002, Narukawa et al 2007).

EMD has partly contrasting effects on bone resorption. The receptor activator of NF-kB ligand

(RANKL) is expressed on the surface of osteoblasts. RANKL’s binding to RANK-receptor on

osteoclast progenitor cells leads to osteoclast differentiation. EMD can reduce the RANKL release

by osteoblasts and enhance the production of osteoblastic osteoprotegerin (OPG), thus inhibiting the

formation and functions of osteoclasts (He et al 2004, Galli et al 2006). Furthermore, the RANKL

mRNA levels of periodontal ligament fibroblast are significantly down-regulated by EMD

(Takayanagi et al 2006). However, it has also been shown that EMD induces osteoclast formation

by interacting with and stimulating the RANKL expressed by osteoblastic cells (Otsuka et al 2005,

21

Itoh et al 2006). In conclusion, EMD may be capable of creating favourable circumstances for bone

regeneration, since it has effects on both bone formation and resorption. Depending on the cell

maturation, EMD can enhance the differentiation of mesenchymal cells into hard tissue-forming

cells and promote their proliferation and growth.

Emdogain and fibroblasts

Several studies have confirmed that EMD enhances the proliferation, migration and in vitro wound

healing of periodontal ligament fibroblasts and gingival fibroblasts (Gestrelius et al 1997,

Lyngstadaas et al 2001, Rincon et al 2003, Cattaneo et al 2003, Davenport et al 2003, Okubo et al

2003, Palioto et al 2004, Keila et al 2004, Rodrigues et al 2007, Zeldich et al 2007). The mitogenic

signalling pathway in which EMD affects PLFs is associated with rapid phosphorylation of the

MAP kinase family and activation of EMD-specific receptor tyrosine kinase (RTK) and

extracellular signal-regulated kinase (ERK) (Kawase et al 2001, Matsuda et al 2002). In addition,

EMD can inhibit gingival fibroblast caspase activation and thereby inhibit TNF-α-induced

apoptosis and enhance survival (Zeldich et al 2007). EMD induces matrix and total protein

synthesis (Gestrelius et al 1997, Haase & Bartold et al 2001, Rodrigues et al 2007). Furthermore,

EMD stimulates insulin-like growth factor (IGF-I), transforming growth factor-β1 (TGF-β1),

platelet-derived growth factor (PDGF) and interleukin-6 (IL-6) production in PLFs (Lyngstadaas et

al 2001, Okubo et al 2003). EMD can increase the production of MMP-2 by PLFs more than

threefold (Mirastschijski et al 2004). In angiogenesis, EMD enhances VEGF release by PLFs

(Mirastschijski et al 2004, Schlueter et al 2007). EMD also has profound effects at the gene level.

EMD can down-regulate genes involved in the early inflammatory phase of wound healing, while

simultaneously up-regulating growth and repair related genes (Parkar et al 2004).

The results from studies on the effects of EMD on the differentiation of fibroblasts are somewhat

contradictory. EMD has been observed to both decrease and increase the ALP activity of PLFs

(Okubo et al 2003, Cattaneo et al 2003, Nagano et al 2004, Rodrigues et al 2007, Lossdörfer et al

2007). In addition, EMD can promote bone-like nodule formation but does not induce osteoblastic

differentiation of PLFs or gingival fibroblasts (Gestrelius et al 1997, Okubo et al 2003, Keila et al

2004, Rodrigues et al 2007). EMD can stimulate differentiation, mineralised tissue formation and

bone metabolism by increasing and modulating OPG and RANKL expression in fibroblasts

22

(Takayanagi et al 2006, Lossdörfer et al 2007). Furthermore, EMD alters phenotype markers of

PLFs when cultured on dental materials (Inoue et al 2004).

2.7 Effects of Emdogain on epithelial cells

There have only been a few studies concerning the effects of EMD on epithelial cell. During

periodontal development, molecules from Hertwig’s epithelial root sheath can induce differentiation

of mesenchymal precursors to form periodontal tissues. It is assumed that EMD can mimic this

process (Slavkin et al 1989, Hammarström et al 1997, Spahr et al 1999, Hakki et al 2001, Zeichner-

David 2001, Gestrelius et al 2004). The proliferation of epithelial cell rests of Malassez (ERM) is

enhanced by EMD (Rincon et al 2005). In addition, EMD can stimulate the expression of

osteopontin by ERM (Rincon et al 2005). In contrast, EMD has no effect on the proliferation of rat

tongue epithelial cells (Gestrelius et al 1997). In vivo, EMD can enhance wound epithelialisation in

rabbits (Mirastschijski et al 2004).

Evidence on the effect of EMD on endothelial cell proliferation is scarce and somewhat

inconsistent. EMD has been demonstrated to either stimulate or to have no effects on the

proliferation of endothelial cells (Yuan et al 2003, Schlueter et al 2007). Both of these studies have

confirmed that EMD increases the chemotaxis of endothelial cells (Yuan et al 2003, Schlueter et al

2007). In vivo mice experiment showed significantly more blood vessels surrounding EMD-treated

collagen implants than controls (Yuan et al 2003). In addition, EMD has been shown to stimulate

the production of MMP-2 by endothelial cells more than threefold (Mirastschijski et al 2004).

2.8 Effects of Emdogain on neoplastic cells

The effects of EMD on malignant cells have been only sparsely studied. Furthermore, most of these

few experiments have been carried out with cells originating outside oral tissues. Thus, these

experiments are not of particular relevance in the context of EMD and oral cancer. These studies

conclude that EMD has either no effects or has an inhibitory effect on the proliferation of malignant

cell lines. No differences in the proliferation of fibrosarcoma HT-1080 cells or breast cancer-

derived MCF-7 cells have been observed after EMD treatment (Gestrelius et al 1997, Takayama et

al 2005). EMD decreases the proliferation of human cervix-derived epithelial HeLa cells, human

osteosarcoma MG-63 cells and human squamous cell carcinoma-derived SCC-25 cells (Kawase et

23

al 2000, Schwartz et al 2000, Lyngstadaas et al 2001). It has also been shown that EMD can arrest

the cell cycle of SCC-25 cells at G1 and do not increase SCC-25 apoptosis (Kawase et al 2000).

Although it has negligible or inhibitory effects on the proliferation of malignant cells, EMD alters

many other cellular functions. A recent study has shown EMD to be capable of enhancing the in

vitro degradation of type I collagen by malignant MG-63 osteosarcoma cells. In addition, EMD was

observed to stimulate the production of MMP-1, but had no effect on the production of MMP-2, -8,

-13, or -14 (Goda et al 2008). Another study using the same MG-63 cells observed that EMD

enhanced in vitro wound fill rates (Hoang et al 2000). However, EMD can inhibit the adhesion of

breast cancer (MCF-7) cells to the bone matrix, but it was concluded that EMD contains BSP-like

molecules (Takayama et al 2005). EMD increases intracellular cAMP levels and PDGF-AB

production of HeLa cells, and the level of TGF-β1 in the culture medium of MG-63 osteosarcoma

cells (Schwartz et al 2000, Lyngstadaas et al 2001). EMD as well as TGF-β1 stimulates

phosphorylation of several MAP family kinases in SCC-25 cells (Kawase et al 2001, 2002).

24

3 A CASE REPORT

A 63-year-old patient was referred to the Institute of Dentistry, Oulu University Central Hospital,

Oulu, Finland by a general dental practitioner for the management of lichenoid mucosal lesions and

periodontal disease. The patient was considered healthy with no obvious risk factors. He had no

medication and was a non-smoker. A clinical examination revealed an almost full dentition with

chronic periodontal disease. Probing depths of 4-5 mm were observed throughout the mouth and

10-12 mm at the buccal aspect of d.12. Radiographs revealed alveolar bone loss distal to d.12 but

no periapical pathology. Ulcerative and reticular lichenoid lesions were localised mostly at the

anterior maxillary gingiva and sulcus but also at the distal aspects of the lower molars. Toluidine

blue stained small spots at the buccal mucosa and sulcus of dd.11-12 and anterior maxillary gingiva.

Incision biopsy from the stained regions was performed and the histological findings confirmed

moderate dysplasia and the mucosal lesions were diagnosed clinically as non-homogenous

leukoplakia.

Figure 2. Prior to treatment, the gingival recession at d.12 and lichenoid lesions on the anterior maxilla are visible.

Figure 3. Periapical radiograph shows alveolar bone loss around and distal to d.12.

25

The periodontal treatment proceeded as planned with an infection control phase by a course of non-

surgical root surface debridement and oral hygiene instructions. A persisting pocket at the

distobuccal aspect of d.12 was surgically managed with EMD. Under local anaesthesia, labial and

palatal full-thickness mucoperiosteal flaps were raised at the region dd. 11-13. The papillary tissue

was preserved to facilitate primary closure and healing. Alveolar bone loss at buccal aspects of dd.

11 - 13 and two-to-three wall defect at the distal aspect extending to the apical third of d.12 were

observed. The defect was mechanically cleaned and prepared for EMD application. The root

surfaces were rinsed with saline, conditioned for three minutes with EDTA and the Emdogain-gel®

applied. The flaps were replaced and sutured. Post-operative instructions included rinsing the area

with 0.2% Corsodyl® for one minute twice a day, and avoiding brushing the area and hard food.

The surgical area was examined after one week and it was observed to be normal with no signs of

infection or other abnormalities. The sutures were removed two weeks post-operative and the area

was normal with minor swelling at the labial aspect.

A month post-operative the area of the surgery was considered normal but the mucosa was

thickened. Candida was diagnosed by Oricult® culture and treated with Daktarin® gel. After two

weeks of treatment the mucosa was subjectively thinner and less uncomfortable. A culture

confirmed the absence of Candida. At the same time, six weeks post-operative, the Corsodyl rinses

were stopped, plaque control performed, and home care instructions repeated. Furthermore, the

treatment of the leukoplakia was started with Bemetson® cortisone cream, applied to custom-made

trays.

After two weeks of cortisone treatment there was no improvement. The medication was continued

for 6 weeks but the response to the treatment was poor. The white mucous lesions with some

ulcerous red areas were spreading. Some plaque was observed but Candida was absent. To rule out

the possibility of allergy interference, the patient was also referred for an allergy test concerning

dental materials. Toluidine blue staining was performed and biopsy samples taken from the stained

regions at the distobuccal aspect of d.12.

The diagnosis from the area was squamous cell carcinoma, exactly four months after the EMD

operation. Three weeks later the carcinoma was operated on. Block resection of anterior alveolar

maxilla from d.11 to d.16 together with the floor of the maxillary sinus was performed. No

metastasis was observed and the margins of the block were free. The partial resection was replaced

26

by a temporary prosthesis and post-operative healing was generally good. Few lichenoid lesions

remained at dorsal aspects of the right maxilla. The lichenoid lesions at the right tuber and also at

the regions of dd. 24 and 37 were later excised with a CO2-laser and no signs of malignancy were

diagnosed.

Figure 4. Clinical appearance of the anterior maxilla shortly after the resection.

After the operations the patient was examined frequently and prosthetic rehabilitation was planned.

One year after the carcinoma resection the patient observed a knot on the anterior maxilla at the

region of d.11. The knot was clinically observed to be verrucosus leukoplakia and the alteration was

excised. The diagnosis was moderate dysplasia. Extended surgery seemed unnecessary. Tree and a

half years later a verrucosus leukoplakia appeared at the buccal region of d.47. Also, gingival

pockets of 5-6 mm were observed buccal to d. 47. The tooth was extracted and a biopsy taken from

surrounding mucosa. The diagnosis was again squamous cell carcinoma, gradus I. Part of the

mandibular bone together with dd. 45, 46 and surrounding mucosa was resected with margins. Two

weeks later the healing was good with little pain and no signs of inflammation. At present, the

patient manages with an upper acrylic partial prosthesis and is re-examined frequently.

In conclusion, the management of lichenoid mucosal lesions and periodontal disease were the aims

of the treatment. Despite the dysplastic lesions, the treatment focused firstly on the periodontal

therapy, and the regenerative surgery was performed in an area with erosive and dysplastic lesions.

A periodontal flap surgery is highly aggravating to tissues as such, and the treatment accompanied

by any regenerative product causes severe irritation to the inflamed tissues. Thus, this treatment was

not a standard procedure and not according to the standard of care. Due to its several limitations,

this case cannot prove the causal role of Emdogain-treatment on the development of oral dysplastic

changes or carcinoma. However, this clinical observation led to the initial hypothesis of the study,

27

i.e. whether EMD could have effects on dysplastic cells in vitro? In addition, this case report also

emphasises the importance of critical evaluation of treatments and treatment protocols as well as the

significance of clinical observations as potential triggers also for in vitro studies.

28

4 HYPOTHESES AND AIMS OF THE STUDY

EMD is currently widely used for treating patients with periodontal disease. The product is said to

be safe with very few systemic or local contraindications (www.straumann.com). However, the

mechanisms of actions and the exact composition of EMD are not completely clear. It has been

shown that EMD can promote regeneration in vivo and the proliferation and growth of several

different cell lines in vitro thus possessing growth factor-like effects. Clinically, EMD is surgically

applied in periodontal pockets where it interacts with proteolytic enzymes, e.g. MMPs, of the

periodontitis affected tissues. Thus, this combination of proteolytic enzymes together with unknown

cytokines of EMD might also be capable of inducing unwanted effects and alterations in vivo.

However, there have been no studies focusing on the associations between EMD and oral dysplasia

or oral proteolytic enzymes. This lack of essential knowledge, which has also clinical relevance, led

to two distinct study hypotheses. The first hypothesis was that the proteases present in oral tissues

and oral fluids can degrade EMD, and the degradation alters the effects of EMD in vitro. The

second hypothesis was that EMD could be capable of inducing carcinogenesis-related factors in

vitro and in vivo. Thus, these hypotheses raised the following aims for each of the four sub-studies:

I. To study the effects of GCF and MMPs on the degradation of EMD. Furthermore, the effects

of GCF-degraded EMD on the proliferation of periodontal ligament fibroblasts were to be

studied and compared to the effects caused by amelogenin.

II. To examine the effects of EMD on migration, proliferation, wound closure, apoptosis,

attachment and gelatinases (MMP-2 and -9) production of highly malignant oral tongue

carcinoma (HSC-3) cells in vitro. In addition, the in vivo effects of EMD on HSC-3 inoculated

mice survival and metastasis were to be studied.

III. To compare the effects of EMD and TGF-β1 on gene expression of HSC-3 cells. In addition,

the study aimed to contribute to the understanding of the composition of EMD by comparing

the effects of EMD and TGF-β1.

IV. To investigate the in vitro effects of EMD on oral tissues and carcinogenesis-related factors

and cytokines by using a systematic literature review.

29

5 MATERIALS AND METHODS

The methods used are described in detail in the original publications and briefly summarised below.

5.1 Patients and sample collection (I)

This study was approved by the Ethics Committee of the Ege University, Izmir, Turkey, and

informed consent, in line with the Helsinki declaration, was obtained from each subject. A total of

80 subjects (35 male, 45 female) were obtained from the Department of Periodontology, School of

Dentistry, Ege University, İzmir, Turkey. The periodontal status of the subjects was determined

according to the clinical and radiographic criteria proposed by the 1999 International World

Workshop (Armitage, 1999). Thus, there were 20 subjects in each of the four intended study groups

i.e. healthy subjects, subjects with gingivitis, subjects with chronic periodontitis and subjects with

acute periodontitis

A total of 156 GCF samples were collected from the subjects, with 39 samples from each group

(healthy, gingivitis, chronic periodontitis and acute periodontitis). Prior to GCF sampling,

supragingival plaque was removed and the surfaces were gently air-dried and isolated with cotton

rolls. GCF samples were collected by inserting filter paper strips into gingival pockets (Periopaper,

ProFlow, Inc., Amityville, NY, USA) for 30 seconds. During this process, care was taken to avoid

mechanical injury, and strips contaminated with blood were discarded. GCF volume was estimated

(Periotron 8000, ProFlow, Inc., Amityville, NY, USA) and strips were placed into a sterile

polypropylene tube prior to freezing at -40°C. The study groups were blinded from the authors who

conducted the in vitro assays.

5.2 Study reagents (I, II, III)

Lyophilised Emdogain® was provided by the manufacturer (Straumann, Basel, Switzerland.

Previously: Biora AB, Malmö, Sweden). The EMD concentrations of 10, 100 and 200 µg/ml used

were selected according to preliminary proliferation and gelatin zymography assays, in which the

effects of EMD at the concentration of 0-400 µg/ml on HSC-3 carcinoma cells were tested (un-

published data not shown). 100 µg/ml was observed to be the lowest concentration giving

significant response, whereas 400 µg/ml did not raised the response compared to 200 µg/ml.

Selected concentrations are also supported by previous publications and they were well below

30

clinical concentrations (30mg/ml; 0,3-0,9ml) (Gestrelius et al 1997, Hoang et al 2000, Haase &

Bartold 2001).

The recombinant porcine amelogenin (rP172) was expressed in E. coli and isolated and

characterised with mass spectrometry as previously described (Ryu et al 1999). EMD is

predominantly amelogenin (>90%; Zeichner-David 2006), thus the amelogenin concentrations used

were equal to the selected EMD concentrations. Recombinant human TGF-β1 was acquired from a

supplier (Sigma-Aldrich, St. Louis, MO). Since the TGF-β1 concentration of EMD is unknown, the

used TGF-β1 concentration of 10 ng/ml was selected based on previous publications (Pääkkönen et

al 2008).

All the study reagents were lyophilised and for the assays dissolved in PBS. To maximise the

dissociation of study reagents to PBS the solutions were gently mixed 24 h in RT.

5.3 EMD degradation assays (I)

Several different EMD concentrations as well as both lyophilised EMD and EMD-gel were first

tested to produce high quality gels. 9 mg/ml EMD-gel embedded in the 11% SDS-PAGE, instead of

the normally-used gelatin (Laemmli 1970), was selected to be used in these studies. To detect

EMD-degrading enzymes in GCF, the GCF samples were dissolved in TNC-buffer (50 mM Tris,

0.2 M NaCl, 1 mM CaCl2, pH 7.8) and the same volume (15 µl/well) of each of the 156 GCF

samples was loaded onto an EMD-embedded zymography gel.

The study also aimed to compare the degradation profile caused by GCF to the degradation profile

caused by MMPs. Therefore, recombinant human MMPs -1, -2, -8, -9, -13 and -14 (Chemicon,

Temecula, CA; Invitek, Berlin, Germany; Calbiochem, San Diego, CA) were loaded (200 ng in 3

µl/well) onto a separate EMD-embedded gel. After electrophoresis, the gels were washed twice

with 2.5% Trixon X-100 buffer to renature the gelatinases. The gels were then incubated in TNC-

buffer 72 h at 37°C, and subsequently stained with 0.5% Coomassie Brilliant Blue R-250. The

intensities of separate bands were scanned and measured quantitatively using optical densitometry

and Quantity One software (Bio Rad Model GS-700 Imaging Densitometer, Bio-Rad, Richmond,

CA, USA). To eliminate the background disturbance of the relatively diffuse bands, the analyses

were done strictly at 97, 50 and 37 kDa and using equal sized boxes in every parallel gel.

31

5.4 Cell lines and cell cultures (I, II, III)

In vitro experiments were carried out using one malignant and two different benign cell lines.

Malignant cells were highly invasive HSC-3 tongue squamous cell carcinoma cells (Japan Health

Science Resources Bank, JRCB 0623). Benign cells were human oral mucosal keratinocytes

(HMKs) and human periodontal ligament fibroblasts (PLFs), which were obtained from surgical

gingival biopsy or by scraping intact premolars as described previously (Mäkelä et al 1999,

Oikarinen & Seppä 1987).

The cells were cultured at 37°C in a humidified atmosphere and in their normal media containing

per 100 ml for HSC-3 cells: 45 ml Dulbecco’s modified Eagle’s medium (DMEM), 45 ml Ham’s

Nutrient Mixture F-12, 10 ml fetal calf serum, 100 units/ml penicillin, 100 µg/ml streptomycin, 50

units/ml nystatin, 250 ng/ml Fungizone®, 1 mM sodium pyruvate, 2 nM L-glutamine and 0.4 ng/ml

hydrocortisone; for HMKs: Keratinocyte-SFM supplement and Supplements For Keratinocyte-

SFM, including human recombinant epidermal growth factor (EGF 1-53) and bovine pituitary

extract, 1 % penicillin and 0.1 % Fungizone®; and for PLFs: 90 ml DMEM, 10 ml fetal calf serum,

100 units/ml penicillin, 100 µg/ml streptomycin and 250 ng/ml Fungizone®; and for PLFs: 90 ml

Dulbecco’s modified Eagle’s medium, 10 ml fetal calf serum, 100 units/ml penicillin, 100 µg/ml

streptomycin, and 250 ng/ml Fungizone® (All supplements excluding hydrocortisone: Life

Technologies, Paisley, Scotland. Hydrocortisone: Diosynth, Oss, Netherlands).

5.5 In vitro cell proliferation, migration, wound closure and adhesion assays (I, II)

Periodontal ligament fibroblast proliferation assay (I)

To determinate in vitro the consequences of the degradation of EMD caused by GCF, the effects of

EMD with or without GCF and also the effects of recombinant amelogenin on the proliferation of

PLFs were studied. The cells were culture in serum containing media on 96-well culture plate at the

density of 5000/well (N=6). Before adding the study reagents, the cells were allowed to attach and

grow overnight. The GCF samples dissolved in TNC-buffer were pooled from all diseased and

healthy samples and sterile filtered before adding 10 µl/well with or without EMD (100 and 200

µg/ml). Recombinant amelogenin (100 and 200 µg/ml) was used without GCF. The control cells

were incubated with or without TNC-buffer (10 µl/well). After 48 h incubation, the proliferation of

32

PLFs was determined with the commercial Cell Proliferation ELISA BrdU kit (Roche Diagnostics,

Mannheim, Germany) according to the manufacturer’s instructions.

Human oral carcinoma cells and benign keratinocytes proliferation assay (II)

The effect of EMD on the proliferation of HSC-3 and HMK cells was studied by culturing the cells

in 96-well plates and in their normal medium for 24 h before adding EMD to give concentrations of

0, 100 or 200 µg/ml. The cells were incubated again for 12, 24, 48, 72 and 96 h followed by the

ELISA proliferation analysis.

Apoptosis (II)

Apoptosis of HSC-3 and HMK cells was determined using Annexin-V-Fluos® (Roche Diagnostics,

Penzberg, Germany) and propidium iodide (Roche Diagnostics, Indianapolis, IN) according to

manufacturer’s instructions.

Transmigration assay (II)

Cell migration was studied using 8.0 µm-pore size and 6.5 µm-diameter Transwell® inserts (Costar,

Cambridge, MA). Before plating, the cells were pre-incubated with 100 or 200 µg/ml EMD for 2 h

at RT. The cells were allowed to migrate overnight and fixed with 10% trichloroacetic acid (TCA).

The cells that had not migrated through the filter were removed and cells on the lower side were

stained with crystal violet and either counted microscopically (HSC-3) or by scanning (HMK) (Bio

Rad Model GS-700 Imaging Densitometer, Bio-Rad, Richmond, CA, US).

To examine whether the increased migration of HSC-3 is due to active MMP-2 or -9, a cyclic

peptide CTT-2 (50 µg/ml) that specifically inhibits gelatinolytic activities of MMP-2 and -9 but not

of other MMPs was added to the transmigration assay with or without EMD (100 or 200 µg/ml)

(Koivunen et al 1999, Heikkilä et al 2006).

In vitro wound closure (II)

Two-well chamber plates were pre-coated with BSA, and then 1x106 HSC-3 and HMKs were

incubated for 1 h in the wells. After pre-incubation the culture plates were wounded with a sterile

33

pipette tip, washed with PBS and exposed to medium containing 0, 10 or 100 µg/ml EMD and

cultured overnight. After incubation, the cultures were fixed with 10 % TCA and stained. The

wounds were scanned and printed, and the wounded area was determined by comparing the weight

of wounds to the control (Bio Rad Model GS-700 Imaging Densitometer, Bio-Rad, Richmond, CA,

USA).

In vitro cell adhesion assay (II)

For the cell adhesion assay, 96-well plates were coated with EMD (10 µg/ml), fibronectin (10

µg/ml, Sigma), BM components containing Matrigel (1:10 in PBS, Becton Dickinson, Bedford,

MA) or BSA (10 µg/ml, Sigma) and left overnight at RT. HSC-3 and HMK cells were then plated

at 10 000 cells/well and incubated for 2 h. The non-adherent cells were rinsed off and the remaining

cells were fixed with 10 % TCA, stained and quantified using an ELISA reader at 540 nm. The

adhesion study was repeated using the same coating components as described above, but prior to

the experiments the cells were pre-incubated 2 h with EMD (100 µg/ml).

5.6 In vitro MMP production assays (II, III, unpublished data)

Gelatin zymography (II)

In order to determine the effects of EMD on gelatinolytic enzymes produced by HSC-3 carcinoma

cells and periodontal fibroblasts, a gelatin zymography of the culture medium was conducted. HSC-

3 cells at a density of 28 000/well and PLFs at 30 000/well were allowed to grow to sub-confluence

in 24-well plates. The HSC-3 medium was replaced with serum-free medium and 0, 100 or 200

µg/ml EMD was added to both assays. Cells were incubated for 48 h and medium collected, briefly

centrifuged and stored at –20°C. The medium was added to 10 % SDS-PAGE containing 1 mg/ml

of fluorescently labeled gelatin (Sutinen et al 1998). After electrophoresis, the gels were washed,

incubated overnight and stained as described above. The intensities of separate bands were

measured quantitatively using optical densitometry and Quantity one software (Bio Rad Model GS-

700 Imaging Densitometer, Bio-Rad, Richmond, CA, USA).

ELISA immunosorbent assay (II)

34

To confirm the effects of EMD on gelatinolytic MMPs observed by zymography, an

immunosorbent assay was done on HSC-3 culture medium. After 48 h of EMD incubation (0, 100

and 200 µg/ml) the MMP-2 and MMP-9 concentrations were analyzed by a commercial ELISA

assay according to the manufacturer's instructions (Biotrak, Amersham Pharmacia Biotech,

Buckinghamshire, UK).

Western immunoblotting (III)

HSC-3 cells at 28 000/well were cultured in serum-free media with EMD (0, 100 or 200 µg/ml) for

24 h. Western blotting for MMP-10 from culture media was carried out as previously described

(Nyberg et al 2003). After SDS-PAGE gel electrophoresis the membrane was incubated with

human monoclonal MMP-10 antibody (1:1000, R&D systems, Minneapolis, MN).

5.7 Affymetrix cDNA microarrays (III)

Microarray sample preparation and hybridization

To assure an adequate amount of RNA, 1x106 HSC-3 cells/flask were seeded and allowed to grow

in their normal media overnight. The study reagents TGF-β1 (10 ng/ml) and EMD (200 µg/ml)

were dissolved from the PBS-stock solutions to serum containing culture media, and the media was

changed. Equal volume of PBS was also added to control cells when the culture media of the cells

was changed. After 6 h and 24 h of incubation the total cellular RNA was extracted using the

Qiagen RNeasy Mini Kit according to the manufacturer’s instructions (Quiagen, Oslo, Norway).

RNA quality and quantity were determined using the Agilent 2100 Bioanalyzer (Agilent

Technologies, DE, USA). Micro-array experiments were carried out using the Affymetrix Human

Genome U133 Plus 2.0 chip, for analysis of over 47,000 transcripts. Total RNA (8 µg) from each

sample was used for target cDNA synthesis according to the Affymetrix protocol

(www.affymetrix.com). Three biological replicates were produced at each time point (Lee &

Whitmore 2002)

Gene Ontology enrichment analysis

For the differentially expressed genes, Gene Ontology (GO) enrichment analysis was performed

using the DAVID annotation tool (http://david.abcc.ncifcrf.gov/home.jsp) according to Huang et al

35

(2009). The analysis was conducted using the list of all statistically significantly up- and down-

regulated genes separately. The whole genome list of the Human Genome U133 Plus 2.0 chip was

used as a background. An EASE score (E-score), which is a modified Fisher Exact P-Value, was

used to identify enriched categories. An E-score of <0.05 was considered statistically significant.

Furthermore, the enriched terms were grouped into functional clusters according to terms having

similar biological meaning due to sharing similar gene members.

5.8 Mouse experiments (II)

The animal experiment was approved by the Ethics Committee for Animal Experimentation in the

State Provincial Office of Southern Finland. For inoculation, the 1x107 detached HSC-3 cells

diluted and suspended in 200 µl serum-free culture medium were injected subcutaneously into the

right and left back of twenty 6- to 8-week old Harlan-Spraque Dawley athymic nude mice. On days

4-8 after the inoculation, ten mice were daily injected subcutaneously with 100 µg/ml EMD in 200

µl saline, whereas ten control mice were daily injected with the same volume (200 µl) of saline.

Three times a week the mice were weighted, the tumour growth measured and the survival

observed. The criteria for euthanasia were tumour size over 10 mm in diameter, weight loss of more

than 15 %, or symptoms of suffering. After euthanasia, the mice were dissected and enlarged

axillary and peritoneal lymph nodes (n=6) were collected for metastasis analysis. The presence of

metastasis in the lymph nodes was confirmed histologically.

5.9 Data collection for the systematic review (IV)

An information specialist-assisted search was made from PubMed and EMBASE databases. The

search terms were Emdogain and/or ‘enamel matrix derivative’ and language definition English.

The publication date was restricted to the period of January 1997 to December 2008. Altogether

403 publications were found. No clinical reports concerning association of Emdogain with oral

carcinoma were found. Thus, the review was restricted to in vitro studies and the criteria for a

systematic review was applied after the in vivo exclusion.

Thus, exclusion criteria were:

1) Article not written in English

2) Clinical studies, e.g. studies dealing with radiology or treatment techniques, outcomes and

surveys.

3) In vivo and in vitro studies concerning histology, immunohistocemistry, etc.

36

4) In vitro studies designed to study bacteria or cells other than oral cells or malignant cells.

5) Review articles and conference reports.

Altogether 76 in vitro studies were included in this study, designed to obtain all available biological

data relevant for carcinoma-related factors and the behaviour of oral cells. Thus, inclusion criteria

were:

1) Articles published in English between 1. January 1997 and 31. December 2008.

2) Assays designed to study human cells present in oral tissues or malignant cells.

3) Assays designed to study cytokines in Emdogain or regulated by Emdogain.

5.10 Statistical analysis (I,II,III)

Analysis of variance models was used to estimate the effects of the experimental substrates on the

various in vitro outcomes. The results are presented as relative comparisons of means between

relevant treatment groups, expressed in terms of fold change or percentage from a given baseline,

together with the 95% confidence intervals (CI). In study I, the CI was calculated with the web-

based Texas University Confidence Interval Calculator

(http://www.stat.tamu.edu/~jhardin/applets/index.html). In study II, the R environment for

statistical computing and graphics was used in computations, especially the function lm() for fitting

linear models (URL:www.R-project.org.).

Reporting statistical significance using P-values has deliberately been avoided where possible.

Instead, statistical certainty/uncertainty has been expressed using 95% confidence intervals, which

are known to be more informative than P-values. When reporting the 95% confidence intervals

(CI), it means that the data is well consistent within these limits. At the same time, the CIs point out

the effect size and error margins. In addition, the ’statistical significance’ can also be judged when

CIs are present. In contrast, P-value is generally uninformative, fails to convey important

information about effect size and is often misleading (lack or over emphasis of the biological

relevance). This statistical method follows also the explicit recommendations given by leading

medical journals (Brennan et al 1994, Sterne et al 2001), and these recommendations are also

adopted in the Uniform Requirements for Manuscripts from 1988 onwards (International

Committee of Medical Journal Editors. Uniform Requirements for Manuscripts Submitted to

Biomedical Journals, URL:www.icmje.org). However, the P-values are presented for the

microarray data and a P-value of <0.05 is considered statistically significant.

37

Microarray data analysis

The microarray data analysis was performed using open source Chipster software

(http://chipster.csc.fi). For normalization, the Robust Multi-Array method was used. The regulated

genes by EMD and TGF-β1 were compared with the gene list of the control cells by using the

empirical Bayes t-test to estimate the statistical significance of individual genes with a cut-off

threshold of P<0.05. The fold change (FC) of the statistically significantly regulated genes was also

calculated.

38

6 RESULTS

6.1 Degradation of EMD by GCF and MMP (I)

GCF was observed to be capable of degrading EMD-embedded zymography gels. The degradation

depended on the state of periodontal disease, with statistically significantly more degradation

observed in all states of periodontal disease (gingivitis, chronic periodontitis and aggressive

periodontitis) compared to healthy control GCF.

To further determine the EMD-degrading proteases observed in periodontic GCF, the effects of

recombinant human MMPs -1, -2, -8, -9, -13 and -14 on EMD- embedded gels were studied. Only

MMPs -2 and -8 caused barely detectable lysis of the gel and in a restricted range compared to GCF

samples.

6.2 In vitro cell proliferation, viability and behaviour assays (I, II)

Effects of EMD, GCF and amelogenin on the proliferation of periodontal ligament fibroblasts

The proliferation of PLFs was markedly induced by EMD (100 µg/ml and 200 µg/ml) after 24 and

48 h of incubations compared with the untreated controls. When the cells were treated with EMD

together with GCF (10 µg/ml; dissolved in TNC-buffer), the induction of the proliferation of PLFs

by EMD was decreased compared to treatment with EMD alone. Thus, EMD at the lower

concentration of 100 µg/ml together with GCF could not induce the proliferation of PLFs. The GCF

alone as well as TNC-buffer alone had no effect on the proliferation of PLFs (data regarding TNC-

buffer not shown). Amelogenin, the most abundant protein in EMD, had a drastic inhibitory effect

on the proliferation of PLFs. On average, 200 µg/ml amelogenin caused a 54% decrease in the

proliferation of PLFs with respect to control.

Effects of EMD on the proliferation and apoptosis of carcinoma cells

In HSC-3 cell proliferation, no differences were found between the control and the EMD-treated

(100 and 200 µg/ml) cells after 12, 24, 48, 72, and 96 h of incubation. Furthermore, EMD (100 and

200 µg/ml) had no effect on apoptosis of HSC-3 or HMK cells.

39

Effects of EMD on in vitro migration and wound closure of carcinoma cells

EMD stimulated the random migration of HSC-3 cells in Transwell-assays during overnight culture.

EMD at 100 µg/ml increased the migration of HSC-3 cells by 43% (95% CI: 21% to 69%) and at

200 µg/ml by 57% (95% CI: 33% to 86%) in relation to control. EMD had no effect on the

migration of benign HMK cells.

CTT-2 (50 µg/ml), a selective inhibitor of MMP-2 and -9, decreased the migration of HSC-3

carcinoma cells by 89% (95% CI: 65% to 97%) compared to the control. In combination with EMD

at 100 µg/ml, CTT-2 decreased the migration by 97 % (95% CI: 91% to 99%), but with EMD at

200 µg/ml the observed migration was decreased to only 57% of the control level (95 % CI: -87%

to +41%). CTT-2 (50 µg/ml) also decreased the migration of benign HMK cells.

EMD was able to induce in vitro wound closure of carcinoma cells, which requires both cell

proliferation and migration. In the beginning of the assay the mean of the wounded areas of HSC-3

cells was 46 mm2 (SD 7.1 mm2). After 24 h incubation the mean areas of control wounds were

reduced to 35.2 mm2 (SD 13.8 mm2). EMD at 10 µg/ml concentration decreased the wound areas

by 51% (95% CI: 10% to 74%), and at 100 µg/ml by about a similar amount: 57% (95% CI: 20% to

77%) compared to the control wounds.

Effects of EMD on carcinoma cell adhesion

EMD was able to alter the attachment of HSC-3 cells to various extracellular matrix components.

The mean attachment of HSC-3 cells was similar for both EMD and BSA coated surfaces. Matrigel

and fibronectin coating increased the attachment of HSC-3 cells relative to BSA by 20% (95% CI:

14% to 26%) and by 8% (95% CI: 4% to 13%), respectively. Pre-incubation with EMD-containing

medium (100 µg/ml) followed by the adhesion experiment increased the HSC-3 adhesion to BSA

by 10% (95% CI: 5% to 16%), to fibronectin by 11% (95% CI: 5% to 16%), but decreased the

adhesion to Matrigel by 7% (95% CI: 2% to 11%), whereas the effect on the adhesion to EMD was

unclear, the observed increase being 3% (95% CI: -2% to +9%)

6.3 Effects of EMD on the in-vitro production of matrix metalloproteinases (II, III, un-

published data)

40

Emdogain-induced production of MMP-2 by periodontal ligament fibroblasts

The production of 72 kDa pro-MMP-2 was observed to be induced by EMD in the conditioned

medium of periodontal ligament fibroblasts when analysed with gelatin zymography. The induced

pro-MMP-2 production was on average 1.7-fold (95% CI: 1.5 to 1.8) with EMD at 100 µg/ml and

1.8-fold (CI 1.6 to 2.0) with EMD at 200 µg/ml with respect to control.

Figure 5. An example of a gelatin zymography gel from the culture medium of periodontal ligament fibroblasts. EMD dose-dependently induced pro-MMP-2 production.

Emdogain-induced production of MMP-2, MMP-9 and MMP-10 by carcinoma cells

The production of 72 kDa pro-MMP-2 and 92 kDa pro-MMP-9 was markedly induced by EMD in

the conditioned medium of HSC-3 carcinoma cells when analyzed with gelatin zymography. The

induction caused by EMD was also stronger in carcinoma cells than in periodontal ligament

fibroblasts. The induced pro-MMP-2 production was on average 2.3-fold (95% CI: 1.6 to 3.2) with

EMD at 100 µg/ml and 2.6-fold (95% CI: 1.8 to 3.7) with EMD at 200 µg/ml with respect to

control. The induced pro-MMP-9 ratios were 2.4-fold (95% CI: 2.0 to 2.8) with EMD at 100 µg/ml

and 2.3-fold (95% CI: 1.9 to 2.7) with EMD at 200 µg/ml with respect to control. The production of

MMP-2 and MMP-9 was similarly induced by EMD in the conditioned medium of HSC-3 cells

analyzed with ELISA.

Western immunoblots showed that the untreated HSC-3 cells did not produce MMP-10 but the

production was clearly induced at both EMD concentrations (100 and 200 µg/ml) after 24 h of

incubation.

6.4 Effects of EMD and TGF-β1 on the gene profile of carcinoma cells (III)

41

Comparison of the genes regulated by TGF-β1 and EMD treatment

The Affymetrix microarray revealed a total of 32 genes to be statistically significantly (P<0.05) up-

regulated and 16 genes down-regulated in HSC-3 carcinoma cells after 6 h of EMD treatment

compared to untreated control cells. After 24 h of EMD incubation, this effect on the regulated

genes (P<0.05) was attenuated, showing only four up-regulated and eight down-regulated genes.

TGF-β1 altered regulation of markedly more genes than EMD after 6 h and 24 h of treatment. A

total of 239 genes were statistically significantly (P<0.05) up-regulated and 154 genes down-

regulated in HSC-3 cells after 6 h of TGF-β1 treatment compared to untreated control cells. After

24 h of TGF-β1 incubation, a total of 242 genes were up-regulated and 104 genes down-regulated

compared to untreated carcinoma cells.

About half (53%) of the genes (n=17) up-regulated by the 6 h EMD treatment were the same as

those up-regulated by the 6 h TGF-β1 treatment. Five (31%) of the genes down-regulated by EMD

were the same as those by TGF-β1. Of the four up- and eight down-regulated genes at 24 h of EMD

treatment, two up- and two down-regulated genes were the same as those affected by the TGF-β1

treatment.

Effects of EMD and TGF-β1 on MMP family genes

MMP-9 was the most significantly (P<0.05) up-regulated of all the genes after 24 h of EMD

treatment as well as after 6 and 24 h of TGF-β1 treatments. MMP-10 was the most up-regulated

gene after 6 h of EMD treatment, but was then the most down-regulated gene after 24 h. In TGF-β1

treated carcinoma cells, MMP-10 was first among the most up-regulated genes after 6 h of

treatment but was then absent at 24 h. Furthermore, MMP-1 was also statistically significantly up-

regulated after 6 and 24 h of TGF-β1 treatments.

Effects of EMD and TGF-β1 on Gene Ontology (GO) groups

GO enrichment analysis revealed several enriched terms after 6 h EMD treatment (Appendix 1).

After grouping the enriched terms into functional clusters, the analysis revealed keratinocyte

differentiation, inflammatory response and proteolysis to be the most up-regulated. Among the

down-regulated genes, the most enriched clusters were cell growth and morphogenesis, circulatory

42

system and cellular developmental processes. Because of the relatively few genes regulated by

EMD after 24 h, the GO enrichment analysis was not performed for that treatment.

GO analysis of 6 h TGF-β1 treatment revealed several enriched terms (Appendix 1), which were

grouped into functional clusters. The most enriched clusters among up-regulated genes were system

development, chemokine activity/inflammatory response, hemopoiesis and angiogenesis. Those

among down-regulated clusters were cell cycle processes, mitosis/M-phase and cell growth. After

24 h TGF-β1 treatment, the most enriched clusters among up-regulated genes were developmental

processes, inflammatory/defence response and angiogenesis. Down-regulated clusters were

regulation of cellular processes, intracellular organelle and calcium homeostasis.

6.5 Effects of EMD on metastasis formation in vivo (II)

The mean number of enlarged lymph nodes containing metastases was 3.1 (range 1 to 5, SD 1.6) in

the EMD group, and 1.5 (SD 0.9) in the NaCl group, the difference being 1.6 (CI 0.2 to 2.8). The

evidence was insufficient to determine whether there were differences in mortality between the two

groups.

6.6 Results of the systematic review (IV)

Altogether 76 studies were found to meet the inclusion criteria, e.g. in vitro studies concerning

cytokines in EMD or regulated by EMD and assays designed to study the effects of EMD on human

oral cells or malignant cells. EMD was found to enhance the in vitro production of several factors

related to carcinogenesis in malignant and benign cell lines. TGF-β1 and MMPs are crucial

regulators in carcinogenesis and their production is enhanced by EMD in many different cell lines

including osteoblasts, fibroblasts and malignant osteosarcoma cells (Schwartz et al 2000,

Mirastschijski 2004, Goda et al 2008). In addition, EMD promotes the production of several other

carcinogenesis-related cytokines by benign cells. These include CTGF, FGF-2, ALP and

osteopontin produced by osteoblasts; IGF-I, PDGF, VEGF, IL-6, ALP and OPG produced by PDL

fibroblasts; and PDGF, VEGF and osteopontin produced by epithelial-derived cells (Hakki et al

2001, Rincon et al 2005). Together with the effects of EMD on production of MMP- and TGF-β1

by malignant cells, EMD also stimulates collagenolysis by osteosarcoma cells (Goda et al 2008).

Furthermore, EMD increases cAMP levels and PDGF-AB production of HeLa cells and stimulates

43

phosphorylation of several MAP family kinases in SCC-25 cells (Lyngstadaas et al 2001, Kawase et

al 2001, 2002). Despite these findings, EMD has been shown to possess negligible or inhibitory

effects on the proliferation of malignant cells and to inhibit breast cancer cell attachment to bone

matrix (Gestrelius et al 1997, Schwartz 2000, Lyngstadaas 2001, Kawase et al 2001, Takayama

2005).

Table 1. In vitro effects of Emdogain on carcinogenesis-related cytokines and cellular processes in malignant and benign cell lines. [+] corresponds to stimulatory, [−] to inhibitory, [0] to unchanged and [X] to non-relevant effects of EMD on the corresponding carcinogenesis-related variable. Each symbol [+], [−] and [0] corresponds to one study, or to one cell line when several cell lines were studied.

44

7 DISCUSSION

7.1 Summary of the main results

The aim of the study was to examine whether gingival crevicular fluid (GCF), which contains

several different proteolytic enzymes, could degrade EMD and alter its biological functions. In

addition, the objective was to study the effects of EMD on carcinogenesis-related factors, in

particular MMPs, using in vitro and in vivo models. This study also aimed to contribute to the

understanding of the composition of EMD.

GCF was observed to degrade EMD depending on the level and severity of the periodontal disease.

The GCF-degraded EMD was less effective in enhancing PLF proliferation than non-degraded

EMD. EMD was observed to enhance the proliferation of PLFs whereas amelogenin inhibited the

proliferation. In contrast, TGF-β1 was more effective and regulated more genes in HSC-3 cells than

EMD. EMD was capable of enhancing in vitro migration and wound closure of HSC-3 cells and to

promote metastasis formation in vivo. In addition, EMD was capable of inducing MMP-2, MMP-9

and MMP-10 production at the protein level and MMP- 9 and MMP-10 expression at the gene

level.

7.2 EMD in periodontal regeneration

The mechanisms of action and the exact composition of Emdogain are not completely clear. In

addition, a systematic Cochrane review suggests that the overall treatment effect of EMD might be

overestimated and that the actual clinical advantages of using EMD are unknown (Esposito et al

2005). Despite the lack of knowledge, EMD is extensively used in regenerative periodontal therapy

in all kinds of patients including medically compromised subjects. Periodontal inflammation is

characterised by increased apical proliferation of gingival sulcular epithelial cells together with

increased activity of proteolytic enzymes in gingival tissue and GCF. Matrix metalloproteinases

(MMPs) and matrix-degrading serine proteinases have been shown to be the main mediators of

pathologic tissue destruction in periodontitis (Sorsa et al 2004, 2006). These proteinases, in

particular MMPs, can degrade most, if not all, extracellular matrix and basement membrane

components of periodontium (Birkedal-Hansen et al 1993, Sorsa et al 2004, 2006).

45

In periodontal treatment, a full-thickness flap procedure is used to apply the EMD in the deepest

periodontal pockets and in contact with periodontally-exposed root surfaces. Detectable amounts of

EMD remain at the site of application on the root surface for two to four weeks after which EMD is

degraded (Gestrelius et al 1997, www.straumann.com). This degradation is presumably due to

proteolytic enzymes present post-operatively and during the healing process in connective tissue

and thus secreted also in GCF.

The tissue breakdown and healing after the periodontal therapy and after the surgical trauma

involve collagen destruction via an MMP-mediated pathway (Birkedal-Hansen 1993, Kinane et al

2003). It has been shown that non-surgical periodontal therapy decreases the MMP levels of GCF

weeks to months after periodontal therapy (Haerian et al 1996, Kinane et al 2003, Marcaccini et al

2010). However, the MMP levels of GCF are not statistically significantly reduced during the

healing period of the first weeks (Kinane et al 2003,) or in sites that do not respond to the treatment

(Mäntylä et al 2003). Furthermore, it has been shown that periodontitis-affected sites continue to

display high levels of MMP-8 and collagenolytic activity also after effective periodontal treatment,

and these levels and activity are in the range of those found in gingivitis (Figueredo et al 2004).

In this study it was observed that GCF can, depending on the periodontal status, degrade EMD.

Clinically EMD is not directly interacting with GCF but with the connective tissue fluids instead.

Furthermore, EMD should only be used after infection control and proper scaling and root planing.

In this assay, GCF was mainly used to mimic the circumstances and proteases present in connective

tissue after periodontal therapy. Thus, the strongest EMD degrading effect caused by GCF from

chronic periodontitis states may be clinically overestimated. However, also healthy GCF was

capable of degrading EMD, thus implicating that connective tissue fluids are capable of degrading

EMD also after the periodontal therapy.

In addition, this study suggest, that MMPs are not the main factors contributing to the observed

degradation, since only MMPs-2 and -8 were capable of degrading EMD, and only to a relatively

low extent. The results suggest that the role of other proteases in GCF are most likely more

important for the degradation of the EMD in vivo. Interestingly, both metalloendoproteinase and

serine protease activities have also been observed in EMD itself (Maycock et al 2002).

The effects of EMD on the proliferation of periodontal ligament fibroblasts have been extensively

studied. Also, the observations of the present study confirmed that EMD can enhance the

46

proliferation of PLFs in vitro. Surprisingly, the recombinant amelogenin protein was observed to

depress the proliferation of PLFs significantly. This observation suggests, as Chong et al (2006)

have previously postulated, that amelogenin itself has very little to do with the proliferation of

periodontal ligament fibroblasts, and that the stimulatory effect of EMD on the proliferation of

PLFs is caused by other factors in EMD. Another novel observation was that EMD together with

GCF was less effective in promoting the proliferation than EMD alone, and with the lower

concentration of EMD (100 µg/ml) the proliferation was not different from that of the control cells.

These results suggest that GCF can degrade or inactivate the growth factor-like molecules in EMD

responsible for proliferation. It should also be noted that due to the high consumption of GCF in

these assays, the GCF samples were pooled, and thus the assay also contained less effective EMD-

degrading GCF from healthy subjects. Thus, this approach is also closer to clinical circumstances

where EMD is not in contact with chronically infected tissues, but healing tissues instead. Taking

these observations together, it can be concluded that the inductive effects of EMD on the

proliferation of PLFs are not caused by amelogenin, and thus the presence of cytokines in EMD is

presumable. In addition, the stimulatory effect of EMD on the proliferation of PLFs, one of the

most crucial steps in periodontal regeneration, may be inhibited if the product comes in contact with

proteolytic enzymes in periodontal tissues also in vivo.

7.3 EMD and carcinogenesis

As stated above, the overall knowledge of the effects of EMD on periodontal tissues is relatively

good. However, many relevant issues, including the cytokine content of EMD, still remain

unresolved. There is evidence suggesting that the epithelial-mesenchymal interactions are important

during periodontal tissue development and regeneration, and that EMD can mimic these processes

(Gestrelius et al 2000). In carcinogenesis, epithelial-mesenchymal interactions act in an

uncontrolled manner and one of the first steps towards malignancy is the epithelial to mesenchymal

transition. After the initial formation of the epithelial malignancy, tumour growth depends on the

proteolytic enzymes and cytokines produced by stromal cells surrounding the tumour. EMD has

been observed to stimulate the production of several cytokines and cellular processes normally

related to periodontal regeneration but also to carcinogenesis (See Table 1).

One of the major cytokines in carcinogenesis is TGF-β1. TGF-β1 stimulates cell growth, MMP

production, angiogenesis, invasion and metastasis in head and neck squamous cell carcinomas

47

(HNSCC) (Derynck et al 2001, Akhurst et al 2001). In addition, TGF-β1 together with other

cytokines, for example those in epidermal growth factor families, can enhance morphological and

invasive phases of the EMT phenotype (Wilkins-Port and Higgins 2007). The present study

examined the effects of TGF-β1 on gene expression of carcinoma cells using Affymetrix

microarrays. TGF-β1 was observed to induce angiogenesis-related genes but to down-regulate

proliferation-related genes in carcinoma cells after 6 h of incubation. However, the inhibitory effect

of TGF-β1 on the proliferation of carcinoma cells was only temporary and was absent after 24 h,

whereas the stimulatory effect of TGF-β1 on angiogenesis remained constant over the observation

period.

EMD has been postulated to contain TGF-β1 (Kawase et al 2001, Suzuki et al 2005) and to enhance

the production of TGF-β1 by several cell lines (Schwartz et al 2000). In the present study, when the

effects of EMD and TGF-β1 on carcinoma cell gene expression were compared, TGF-β1 was

observed to regulate markedly more genes than EMD. This result suggests that EMD might not

contain TGF-β1, or that the cytokine is present in other concentrations as studied here. The latter

assumption is supported by the result showing, that the expression of TGF-β family genes was

down-regulated by EMD. Thus, EMD may contain TGF-β1 or have TGF-β1-like activity. In

addition, several regulatory functions for both of the study reagents were similar, for example both

EMD and TGF-β1 up-regulated the expression of MMP-9 and MMP-10 in carcinoma cells. Thus,

this study cannot confirm or reject the hypothesis that EMD contains TGF-β1.

Together with TGF-β1, MMPs are considered to be crucial in tumorigenesis (Lόpez-Otίn &

Matrisian 2007). Typically, several MMPs are expressed in human malignant tumours, and elevated

MMP levels correlate with aggressive and invasive tumours and poor prognosis (Stetler-Stevenson

& Yu 2001). MMPs have an important role at all stages of tumorigenesis: they enhance tumour-

induced angiogenesis, regulate and activate growth factors, and break down ECM and BM to allow

tumour cell invasion and metastatic spread (McCawley & Matrisian 2000; Liotta and Kohn 2001;

Stetler-Stevenson & Yu 2001, Nyberg et al 2008). In this study EMD was observed to enhance the

production of MMP-2 and MMP-9 by carcinoma cells as well as by non-malignant keratinocytes.

The MMPs from the culture media of different assays were analysed both by gelatin zymography

and ELISA immunosorbent assay. The present study confirmed EMD-induced up-regulation of

MMPs at the gene level and the production of MMP-10 in protein level. In addition, the GO

enrichment analysis revealed the term proteolysis being one of the most up-regulated by carcinoma

48

cells after EMD incubation. Also, EMD-induced migration of carcinoma cells was observed to

depend on MMP-2 and -9, since an inhibitor of these MMPs (CTT-2) arrested the effects of EMD

on the migration. Thus, these studies provide extensive results showing EMD to be capable of

regulating MMPs produced by oral carcinoma cells. In addition, other studies have confirmed that

EMD can enhance the production of MMPs in different cell lines. Goda et al (2008) observed EMD

to enhance the production of MMP-1 in malignant MG-63 osteosarcoma cells and to promote the in

vitro degradation of type I collagen. In benign cells, it has been shown that EMD increases more

that threefold the release of MMP-2 from cultured rat skin fibroblasts as well as from endothelial

cells (Mirastschijski et al 2004). This effect of EMD on MMPs is highly interesting and may

potentially cause carcinogenic effects. However, the effects of EMD on MMP production can also

vary between cell lines, MMPs studied and according to other factors and cytokines present in vivo.

As with carcinoma cells, EMD was also observed to enhance the MMP-2 production by periodontal

ligament fibroblasts in vitro. However, EMD decreases the levels of MMP-1 and MMP-8 in GCF

after flap surgery in vivo. In that study, the decreased MMP levels in GCF may just be due to the

decreased inflammation in periodontal tissues after periodontal treatment. Even then the result

highlights the limits of in vitro studies and the variability in the response to different cytokines by

different cell lines and tissues.

As with the production of MMPs, the effects of EMD on other cellular functions also vary between

the cell lines investigated. In this study, notably increased in vitro migration and wound closure of

carcinoma cells were observed after EMD treatment. The effect of EMD on the migration and

wound closure of benign keratinocytes (HMK cells) was less obvious. These findings with a shorter

culture period are similar to previous findings in which enhanced wound-fill of osteosarcoma (MG-

63) cells was observed after a longer culture period (6 to 9 days) (Hoang et al 2000). According to

the effects of EMD on migration and wound closure of carcinoma cells, it would be logical to find

EMD to enhance the proliferation of carcinoma cells as well. However, it has been demonstrated

that EMD has no effect on highly malignant MCF-7 breast carcinoma cell proliferation during short

180 min incubations (Takayama et al 2005). The findings of the present study also demonstrated

that EMD exhibits no effect on the proliferation of oral carcinoma cells after 12h to four days of

treatment. In contrast, EMD inhibited HMK cell proliferation after 12 and 24 hours incubation.

Kawase et al (2001) have shown that EMD inhibits the proliferation of human squamous cell

carcinoma SCC-25 cells after three days of treatment. In the same study, it was also shown that

EMD dose-dependently arrested the cell cycle of SCC-25 cells at G1 and did not discernibly

increase SCC-25 apoptosis over 8 days of treatment. These results indicate that the effects of EMD

49

on cell proliferation are likely to depend on the cell type investigated. The discrepancy between the

results with SCC-25 cells of Kawase et al (2001) and those with HSC-3 cells used in this study may

reflect the fact that HSC-3 cells are highly invasive whereas the SCC-25 cells are moderately

differentiated, less invasive and often, like in the study by Kawase et al (2001), used as a substitute

for normal epithelial cells (Matsumoto et al 1989). Eventually, the less invasive carcinoma cells are

likely to be more sensitive to the regulatory cell proliferation signals of EMD relative to more

invasive and aggressive counterparts.

After having observed carcinogenesis-related factors to be regulated in several in vitro assays, this

study aimed to examine the effects of EMD on carcinogenesis in vivo. Despite the small number of

study animals, ten subjects in the study group and ten in the control group, the inductive effect of

EMD on metastasis was clear, and also statistically significant. Due to the small number of subjects

no differences were observed in survival or tumour growth between the two study groups. Thus, a

more extensive in vivo study would be needed to confirm the observed effects.

In addition to the present study, only a few studies concerning EMD and malignant cells have been

conducted. None of these have focused on carcinogenesis as such, but malignant cells have rather

been used as an “easy growing” substitute for their benign precursors. Also, many of these assays

have been carried out with cells very different from those in oral tissues. Thus, these studies are not

of major relevance in the context of EMD and oral cancer. No matter their original scientific goals,

the present study reviewed all in vitro studies in which malignant cells were treated with EMD.

According to the reviewed studies, EMD has been shown to stimulate the production of several

cytokines known to be up-regulated in carcinomas, including MMP-1, -2, -9, IL-6, TGF-β1 and

PDGF. In addition, EMD can enhance the expression of these and several other cytokines by benign

cells. In carcinogenesis, the cytokines produced by stromal cells surrounding the malignant tissue

are crucial in regulating tumour behaviour and growth. Thus, in a situation in which EMD were

present in direct contact with either malignant tissue or nearby normal tissue, EMD could be

capable of promoting carcinogenesis by inducing cytokine production and transformation of the

tissues.

The initial hypothesis for the present study arose from a clinical observation in which a dysplastic

lesion was converted into an invasive carcinoma shortly after EMD treatment. The presented

clinical situation is rare, but it should be noted that according to the manufacturer of EMD

50

(Straumann, Basel, Switzerland), mucosal dysplasia is not a contraindication for EMD treatment.

However, the treatment choices should not be based solely on the statements made by

manufacturers, and performing a periodontal surgery to an area with an erosive and dysplastic

lesions is not in accordance with the standard of care. In this case, the treatment should have

included the management of the periodontal pocket by conventional scaling and root planing,

together with the treatment of the mucous dysplastic lesions by local medication or by surgical

means. In addition, when considering the possible causal role of Emdogain in the development of

carcinoma, it should be noted that preceding factors existed, as the patient already had widespread

dysplastic pre-malignant non-homogenous leukoplakia lesions. Before the EMD-therapy, several

biopsies were taken around the mucous membranes, and also near the EMD-treated sites, showing

moderate dysplasia. However, it is well known that the degree of dysplasia can vary from site to site

and thus it may be possible that the patient already had a carcinoma in situ. Nonetheless, the

location of the initial cancer within the bone, close to the apex of d.12, and the close chronological

and spatial relationship to the EMD-treatment were unusual. Based on the follow-up history, it is

highly possible that this patient had genetic susceptibility to dysplastic transition (“field

cancerization”), since the recurrences of mucous dysplastic and cancer lesions have occurred

several times during the recent past. However, the possibility of EMD influence on the cancer

induction of this patient could not be completely ruled out. Further studies of the effects of EMD on

cytokines produced by benign and malignant oral tissues together with studies focusing on the

composition of EMD are needed before the actual safety of EMD on patient with dysplastic

mucosal lesions can be confirmed.

7.4 Methodological considerations

The first sub-study examined whether the degradation of EMD caused by GCF has effects on the

proliferation of PLFs. The proliferation was determined with a commercial Cell Proliferation

ELISA BrdU kit. BrdU is incorporated into newly synthesised DNA, which is optically measured.

Thus, the assay actually measures the DNA synthesis of the cells assumed to reflect the

proliferation instead of directly measuring the proliferation of the cells. However, the method was

selected because it is simple and sensitive, does not rely on microscopical counting of the cells, and

is therefore objective and currently often used.

The main result of the proliferation assay was that GCF together with EMD cannot stimulate the

proliferation of PLFs, as EMD did alone. In addition, the GCF treatment alone had no effect on the

51

proliferation. During the proliferation assay, the cells were kept in serum containing media to

simulate in vivo circumstances, in which the cells normally have enough nourishment. However, the

high concentrations of serum in the media are likely to override the effects of tiny amounts of

cytokines and other effectors in GCF. Thus, the result concerning the GCF only treated control cells

may be interfered by the serum. However, this problem cannot cause bias to the main result of the

study, as it was shown that GCF can degrade EMD or inactivate its effects despite the presence of

serum.

The general aim of the second and third sub-studies was to study whether EMD could induce

carcinogenesis. This was studied with in vitro cell assays and in vivo mice experiment. With the

selected in vitro assays, several crucial cellular processes related to carcinogenesis could be studied,

including migration, proliferation and MMP production. However, carcinogenesis relies on several

other factors as well which were not studied. These include the production of other MMPs and

growth factors as well as the invasion capabilities of the cells. In addition, it has been shown that

malignant cells with different phenotypes behave differently in various assays (Matsumoto et al

1989). However, the assays in the present study were done using only one highly malignant cell

line, the HSC-3 oral carcinoma cells. Furthermore, the selected in vitro assays were relatively

simple and thus are not comparable with the in vivo circumstances in which the malignant cells are

constantly interacting with the stromal cells and numerous different cytokines. Thus, the results

from these in vitro studies are presumably partly cell type-specific, and cannot be directly

generalised to all dysplastic cell types or to in vivo circumstances.

In the present assays, EMD, amelogenin and TGF-β1 were dissolved in PBS. In previous studies,

EMD has most often been dissolved in acidic acid (Hoang et al 2000), trifluoroacetic acid (Kawase

et al 2000) or propylene glycol alginate (PGA, Yuan et al 2003) instead of PBS (Klein et al 2007).

In the study, PBS was chosen because it is neutral and thus closer to the neutral pH of the oral

cavity compared to the acids. However, enamel matrix proteins, in particular amelogenin, are

hydrophobic and do not completely dissolve at neutral pH. To maximise the dissociation of EMD to

PBS the solution was gently mixed 24 h in RT. Microscopically, EMD was observed to be properly

dissolved in the culture media after incubation at +37°C. However, there is the possibility that some

minority of the insoluble proteins may not have dissolved and were thus not present in the assays.

This could cause slightly higher concentrations of the possible water-soluble growth factors in

EMD in assays compared to studies in which acid vehicles were used.

52

The used EMD and amelogenin concentrations were selected based on preliminary assays and/or

previous studies. In the case of EMD, the selected concentrations are most likely well below the

concentrations used clinically, since gel-like EMD is applied locally, whereas such gel application

and high concentrations are not possible in cell cultures. This discrepancy between clinical and the

study concentration causes bias, and the observed effects can be highly concentration depend.

However, since clinical concentrations are higher than the used study concentrations, it is likely that

the proteins and possible cytokines of EMD diffuse into the surrounding tissues in vivo, thus

resulting in the study concentrations at some point near the EMD treated site.

Also the selected TGF-β1 concentration (10 ng/ml) in microarrays is somewhat contradictory. The

used concentration was selected to be biologically meaningful according to previous studies

(Pääkkönen et al 2008), while it is unclear whether EMD contains TGF-β1 or not and in what

concentration. However, the response of carcinoma cells to growth factors and to TGF-β1 can be

variable depending on the concentration. Thus, the current microarray analysis can only prove that

EMD cannot contain TGF-β1 at the concentration of 10 ng/ml, but other concentrations of TGF-β1

in EMD are still possible which warrants further studies.

In addition to TGF-β1 concentration, also the data analysis of microarray results had limitations.

Due to the lack of more sophisticated statistical methods available in the used microarray analysis

software, and in microarray analysis software in general, the statistical significance in gene

expression between study and control groups were judged according to P-values instead of CIs,

which were otherwise used in this study. Also the fold change (FC) of the genes between control

and study groups was calculated, but all the statistically significant results were regarded as true

observations with biological relevance. This may cause overstatements in the results since

′statistically significantly′ differing genes can have only relatively small difference in fold change at

which point the biological relevance of the observation may be questionable.

The fourth sub-study aimed to systematically review the existing studies focusing on the effects of

EMD on carcinogenesis-related factors in vitro. There were no previous studies directly focusing on

the associations among oral carcinoma cells and EMD. Therefore, the aim of the study was met by

including all in vitro studies relevant to carcinoma-related factors and behaviour of oral cells. With

the chosen method, the study was able to present and summarise all available biological data

53

concerning the topic. However, at the same time, the inclusion/exclusion criteria had to be relatively

imprecise and it did not thus completely meet the criteria of systematic reviews. First and foremost,

the inclusion/exclusion of the studies was not done according to the methodological diversities of

the reviewed studies. Thus, in the included studies, different experimental set-ups were used

including different cell lines, assay designs, EMD concentrations, dissolvent of EMD, time points,

and controls. However, the underlying biological mechanism, and thus the results of a study, may

depend on the assay design. Therefore, the conclusions of the review must be made taking these

limitations into account.

54

8 CONCLUSIONS

This study aimed to find out the effects of proteolytic enzymes present in GCF on the degradation

and functions of EMD. In addition, the objective was to find out whether EMD could induce

carcinogenesis-related genes, proteins and cellular processes in vitro or in vivo. Furthermore, the

study aimed to contribute to the understanding of the composition of EMD, which has been

observed to contain mainly amelogenin and possibly TGF-β1. The results of this study led to the

following conclusions:

I. GCF containing various proteases degrades EMD. EMD degradation depends on the state of

periodontal disease being stronger in all states of periodontal disease (gingivitis, chronic and

acute periodontal) compared to a healthy state. MMPs are probably not responsible of the

degradation of EMD.

II. GCF can decrease EMD induced proliferation of periodontal ligament fibroblasts. Unlike

EMD, also amelogenin decreases the migration of PLFs. Thus, the effect of EMD on the

proliferation is likely to be mediated by other cytokines in it and not the amelogenin itself.

III. TGF-β1 regulates more and mostly different genes than EMD. This suggests that TGF-β1

cannot be present in EMD at the studied concentration, if present at all.

IV. EMD can induce several factors closely related to carcinogenesis in gene, protein, cell and in

vivo levels. EMD enhance the expression of MMP-10 and MMP-9 in gene level and the

production of gelatinases MMP-2 and MMP-9 and MMP-10 by cultured carcinoma cells.

EMD stimulates the migration and in vitro wound closure of carcinoma cells. Furthermore,

EMD also promotes metastasis formation in athymic mice bearing human tongue carcinoma

xenografts. These results show that EMD can regulate several different factors, not only in

benign cells, but also in highly malignant oral carcinoma cells. These results may have clinical

implications and warrant further studies.

V. Review of the literature confirms that EMD can regulate the production of several cytokines

related to carcinogenesis in benign and malignant cell lines, but the evidence is scarce and in

part contrasting.

55

ACKNOWLEDGEMENTS

This study was carried out at the Department of Cell Biology and Oral Diseases, Biomedicum

Helsinki, Institute of Dentistry, University of Helsinki and at the Department of Diagnostics and

Oral Medicine, Institute of Dentistry, University of Oulu, during the years 2004-2010. Animal

experiments were performed at the Viikki Laboratory Animal Center of the University of Helsinki.

The study was financially supported by the HUCH-EVO and OUCH-EVO grants and by the

Finnish Dental Society Apollonia.

I have been very fortune to have had Professor Tuula Salo and Professor Timo Sorsa as my

supervisors. Their unbeatable team - full of enthusiasm, ideas and passion for science - has inspired,

guided and encouraged me during these years. Thank you.

I wish to express my gratitude to the former and present Deans of the Institute of Dentistry,

University of Helsinki, Professor Jukka H. Meurman and Professor Jarkko Hietanen, for providing

the excellent research facilities at my disposal. I am grateful to Professor Lari Häkkinen and Dr.

Marilena Vered for reviewing my thesis and offering constructive suggestions to improve the work.

I warmly thank all my co-authors for their collaboration: Professor Esa Läärä for his expertise in the

statistics, Dr. Saynur Vardar-Sengul, Professor Gül Atilla, Dr. Buket Han Saygan, Dr. Haluk Baylas

and Professor James P. Simmer for providing the study reagents and samples at my disposal, and

Dr. Juho Suojanen, Sini Nurmenniemi, MSc and Suvi-Tuuli Vilen, DDS, for their participation in

the study assays. I am also thankful to Kirsti Kari, MSc, the Head of the Scientific Laboratory of the

Institute of Dentistry, University of Helsinki, for her organisational skills and kind help with

anything needed in the lab, and to Docent Taina Tervahartiala for her practical support and

guidance. I acknowledge Ritva Keva, Jukka Inkeri, and Marjatta Kivekäs (University of Helsinki)

as well as Maija-Leena Lehtonen (University of Oulu) for their excellent technical assistance. I

thank the entire staff, colleagues and friends at our laboratory for the enjoyable and encouraging

working atmosphere. Outside the lab, I also wish to thank Professor Tuomas Waltimo (University

of Basel), Roope Lehtinen, DDS, and Pirjo Rissa, DDS, for their flexibility, support and

understanding towards my work on the thesis during these years.

My dearest thanks go to my friends and family for their support and for providing me with activities

and thoughts unrelated to science. Finally, words fail me to express my appreciation to my wife

Elina. She has been my third supervisor, the unnamed co-author and the most valuable reviewer of

56

this work at its every phase. She has shared the moments of joy and encouraged me in

disappointments. Thank you for everything.

Basel, April 2010

Matti Laaksonen

57

REFERENCES: Ahn GO, Brown JM. Matrix metalloproteinase-9 is required for tumor vasculogenesis but not for angiogenesis: role of bone marrow-derived myelomonocytic cells. Cancer Cell 2008;13:193–205. Akhurst RJ, Derynck R. TGF-beta signaling in cancer--a double-edged sword. Trends Cell Biol 2001;11:44-51. Armitage GC. Development of a classification system for periodontal diseases and conditions. Ann Periodontol 1999;4:1-6. Armitage GC. Analysis of gingival crevice fluid and risk progression of periodontitis. Periodontol 2000 2004;34:109-119. Aung PP, Oue N, Mitani Y et al. Systematic search for gastric cancer-specific genes based on SAGE data: melanoma inhibitory activity and matrix metalloproteinase-10 are novel prognostic factors in patients with gastric cancer. Oncogene 2006;25:2546-2557. Baruch RR, Melinscak H, Lo J, Liu Y. Yeung O. Hurta RA. Altered matrix metalloproteinase expression associated with oncogene-mediated cellular transformation and metastasis formation. Cell Biol Int 2001;25:411-420. Bendrik C, Robertson J, Gauldie J, Dabrosin C. Gene transfer of matrix metalloproteinase-9 induces tumor regression of beast cancer in vivo. Cancer Res 2008;68:3405–3412. Bergers G, Brekken R, McMahon G, et al. Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nature Cell Biol 2000;2:737–744. Birkedal-Hansen B, Pavelic ZP, Gluckman JL, Stambrook P, Li YQ, Stetler-Stevenson WG. MMP and TIMP gene expression in head and neck squamous cell carcinomas and adjacent tissues. Oral Dis 2000;6:376-382. Birkedal-Hansen H. Role of matrix metalloproteinases in human periodontal diseases. J Periodontol 1993;64:474-484. Birkedal-Hansen H, Moore WG, Bodden MK et al. Matrix metalloproteinases: a review. Crit Rev Oral Biol Med 1993;4:197-250. Björklund M, Koivunen E. Gelatinase-mediated migrtation and invasion of cancer Bryostatin 1/ionomycin (B/I) ex vivo stimulation preferentially activates L-selectinlow tumor-sensitized lymphocytes. Int Immunol 2004;16:1283-1294. Björklund M, Koivunen E. Gelatinase-mediated migratation and invasion of cancer cells. Biochem Biophys Acta 2005;1755:37–69. Bosshardt DD. Biological mediators and periodontal regeneration: a review of enamel matrix proteins at the cellular and molecular levels. J Clin Periodontol 2008;35:87-105. Brennan P, Croft P. Interpreting the results of observational research: chance is not such a fine thing. BMJ 1994;309:727-730 Brinckerhoff CE, Matrisian LM. Matrix metalloproteinases: a tail of a frog that became a prince. Nat Rev Mol Cell Biol 2002;3:207–214.

58

Carinci F, Piattelli A, Guida L et al. Effects of Emdogain on osteoblast gene expression. Oral Dis 2006;12:329-342. Cattaneo V, Rota C, Silvestri M et al. Effect of enamel matrix derivative on human periodontal fibroblasts: proliferation, morphology and root surface colonization. An in vitro study. J Periodontal Res 2003;38:568-574. Cawston TE, Wilson AJ. Understanding the role of tissue degrading enzymes and their inhibitors in development and disease. Best Pract Res Clin Rheumatol 2006;20:983–1002. Chin CS, Miller CH, Graham L et al. Bryostatin 1/ionomycin (B/I) ex vivo stimulation preferentially activates L-selectinlow tumor-sensitized lymphocytes. Int Immunol 2004;16:1283-1294. Chin D, Boyle GM, Theile DR, Parsons PG, Coman WB. Molecular introduction to head and neck cancer (HNSCC) carcinogenesis. Br J Plast Surg 2004;57:595-602. Chin D, Boyle GM, Porceddu S, Theile DR, Parsons PG, Coman WB. Head and neck cancer: past, present and future. Expert Rev Anticancer Ther 2006;6:1111-1118. Chong CH, Carnes DL, Moritz AJ, et al. Human periodontal fibroblast response to enamel matrix derivative, amelogenin, and platelet-derived growth factor-BB. J Periodontol 2006;77:1242-1252. Cimasoni G. Crevicuar fluid update. In Myers H (Ed.). Monographs in Oral Science vol 12 Basel: Karger,1983:45-102. Cornelius LA, Nehring LC, Harding E et al. Matrix metalloproteinases generate angiostatin: effects on neovascularization. J Immunol 1998;161:6845-6852. Cui W, Fowlis DJ, Bryson S et al. TGFbeta1 inhibits the formation of benign skin tumors, but enhances progression to invasive spindle carcinomas in transgenic mice. Cell 1996;86:531-542. Curtis MA, Sterne JA, Price SJ et al. The protein composition of gingival crevicular fluid sampled from male adolescents with no destructive periodontitis: baseline data of a longitudinal study. J Periodontal Res 1990;25:6-16. Davenport DR, Mailhot JM, Wataha JC, Billman MA, Sharawy MM, Shrout MK. Effects of enamel matrix protein application on the viability, proliferation, and attachment of human periodontal ligament fibroblasts to diseased root surfaces in vitro. J Clin Periodontol 2003;30:125-131. Dean DD, Lohmann CH, Sylvia VL et al. Effect of porcine fetal enamel matrix derivative on chondrocyte proliferation, differentiation, and local factor production is dependent on cell maturation state. Cells Tissues Organs 2002;171:117-127. Derynck R, Akhurst RJ, Balmain A. TGF-beta signaling in tumor suppression and cancer progression. Nat Genet 2001;29:117-129. Devarajan P, Johnston JJ, Ginsberg SS, Van Wart HE, Berliner N. Structure and expression of neutrophil gelatinase cDNA. Identity with type IV collagenase from HT1080 cells. J Biol Chem 1992;267:25228-25232. Ding Y, Uitto VJ, Firth J, et al. Modulation of host matrix metalloproteinases by bacterial virulence factors relevant in human periodontal diseases. Oral Dis 1995;1:279-286. van den Dolder J, Vloon AP, Jansen JA. The effect of Emdogain on the growth and differentiation of rat bone marrow cells. J Periodontal Res 2006;41:471-476.

59

Douglas WG, Tracy E, Tan D, et al. Development of head and neck squamous cell carcinoma is associated with altered cytokine responsiveness. Mol Cancer Res 2004;2:585–593. Duffy MJ, Maguire TM, Hill A, McDermott E, O'Higgins N. Metalloproteinases: role in breast carcinogenesis, invasion and metastasis. Breast Cancer Res 2000;2:252–257. Egeblad M, Werb Z, New functions for the matrix metalloproteinases in cancer progression. Nat Rev Cancer 2002;2:161–174. Eley BM, Cox SW. Proteolytic and hydrolytic enzymes from putative periodontal pathogens: characterization, molecular genetics, effects on host defenses and tissues and detection in gingival crevice fluid. Periodontol 2000 2003;31:105-124. Esposito M, Coulthard P, Thomsen P, Worthington HV. Enamel matrix derivative for periodontal tissue regeneration in treatment of intrabony defects: a Cochrane systematic review. J Dent Educ 2004;68:834-844. Esposito M, Grusovin MG, Coulthard P, Worthington HV. Enamel matrix derivative (Emdogain) for periodontal tissue regeneration in intrabony defects. Cochrane Database Syst Rev 2005;19:CD003875. Fidler IJ. Critical factors in the biology of human cancer metastasis. Am Surg 1995;61:1065-1066. Figueredo CM, Areas A, Miranda LA, Fischer RG, Gustafsson A.The short-term effectiveness of non-surgical treatment in reducing protease activity in gingival crevicular fluid from chronic periodontitis patients. J Clin Periodontol 2004;31:615-619. Fiore E, Fusco C, Romero P, Stamenkovic I. Matrix metalloproteinase 9 (MMP-9/gelatinase B) proteolytically cleaves ICAM-1 and participates in tumor cell resistance to natural killer cell-mediated cytotoxicity. Oncogene 2002;21:5213-5223. Folkman J. Angiogenesis. Annu Rev Med 2006;57:1–18. Folkaman J. Angiogenesis: an organizing principle for drug discovery? Nature Rev Drug Discovery 2007;6:273–286. Franchi A, Santucci M, Masini E, Sardi I, Paglierani M, Gallo O. Expression of matrix metalloproteinase 1, matrix metalloproteinase 2, and matrix metalloproteinase 9 in carcinoma of the head and neck. Cancer 2002;95:1902-1910. Friedl P, Wolf K. Tumor-cell invasion and migration: diversity and escape mechanisms. Nature Rev Cancer 2003;3:362–74. Frisch SM, Morisaki JH. Positive and negative transcriptional elements of the human type IV collagenase gene. Mol Cell Biol 1990;10:6524-6532. Galli C, Macaluso GM, Guizzardi S, Vescovini R, Passeri M, Passeri G. Osteoprotegerin and receptor activator of nuclear factor-kappa B ligand modulation by enamel matrix derivative in human alveolar osteoblasts. J Periodontol 2006;77:1223-1228. Gestrelius S, Andersson C, Johansson AC, et al. Formulation of enamel matrix derivative for surface coating. Kinetics and cell colonization. J Clin Periodontol 1997;24:678-684. Gestrelius S, Andersson C, Lidstrom D, Hammarstrom L, Somerman M. In vitro studies on periodontal ligament cells and enamel matrix derivative. J Clin Periodontol 1997;24:685-692.

60

Gestrelius S, Lyngstadaas SP, Hammarström L. Emdogain--periodontal regeneration based on biomimicry. Clin Oral Investig 2000;4:120-125. Giambernardi TA, Grant GM, Taylor GP, et al. Overview of matrix metalloproteinase expression in cultured human cells. Matrix Biol 1998;16:483-496. Gill JH, Kirwan IG, Seargent JM et al. MMP-10 is overexpressed, proteolytically active, and a potential target for therapeutic intervention in human lung carcinomas. Neoplasia 2004;6:777-785. Giraudo E, Innoue M, Hanahan D. An amino-bisposphonate targets MMP-9 expressing macrophages and angiogenesis to impair cervical carcinogenesis. J Clin Invest 2004;114:623–633. Goda S, Inoue H, Kaneshita Y et al. Emdogain stimulates matrix degradation by osteoblasts. J Dent Res 2008;87:782-787. Gottlow J, Nyman S, Karring T, Lindhe J. New attachment formation as the result of controlled tissue regeneration. J Clin Periodontol 1984;11:494-503. Graesslin O, Cortez A, Fauvet R, Lorenzato M, Birembaut P, Darai E. Metalloproteinase-2, -7 and -9 and tissue inhibitor of metalloproteinase-1 and -2 expression in normal, hyperplastic and neoplastic endometrium: a clinical- pathological correlation study. Ann Oncol 2006;17:637–645. Guida L, Annunziata M, Carinci F, Di Feo A, Passaro I, Oliva A. In vitro biologic response of human bone marrow stromal cells to enamel matrix derivative. J Periodontol 2007;78:2190-2196. Gurpinar A, Onur MA, Cehreli ZC, Tasman F. Effect of enamel matrix derivative on mouse fibroblasts and marrow stromal osteoblasts. J Biomater Appl 2003;18:25-33. Görögh T, Beier UH, Bäumken J, Meyer JE, Hoffmann M, Gottschlich S, Maune S. Metalloproteinases and their inhibitors: influence on tumor invasiveness and metastasis formation in head and neck squamous cell carcinomas. Head Neck 2006;28:31-39. Haase HR, Bartold PM. Enamel matrix derivative induces matrix synthesis by cultured human periodontal fibroblast cells. J Periodontol 2001;72:341-348. Hagedorn H, Sauer U, Schleicher E, Nerlich A. Expression of TGF-beta 1 protein and mRNA and the effect on the tissue remodeling in laryngeal carcinomas. Anticancer Res 1999;19:4265-4272. Hakki SS, Berry JE, Somerman MJ. The effect of enamel matrix protein derivative on follicle cells in vitro. J Periodontol 2001;72:679-687. Hammarström L. Enamel matrix, cementum development and regeneration. J Clin Periodontol 1997;24:658-668. Hammarström L, Heijl L, Gestrelius S. Periodontal regeneration in a buccal dehiscence model in monkeys after application of enamel matrix proteins. J Clin Periodontol 1997;24:669-677. Hanahan D, Folkman J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 1996;86:353-364. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57-70. He J, Jiang J, Safavi KE, Spångberg LS, Zhu Q. Direct contact between enamel matrix derivative (EMD) and osteoblasts is not required for EMD-induced cell proliferation. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004;98:370-375.

61

He J, Jiang J, Safavi KE, Spångberg LS, Zhu Q. Emdogain promotes osteoblast proliferation and differentiation and stimulates osteoprotegerin expression. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004;97:239-245. He J, King Y, Jiang J, Safavi KE, Spångberg LS, Zhu Q. Enamel matrix derivative inhibits TNF-alpha-induced apoptosis in osteoblastic MC3T3-E1 cells. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2005;99:761-767. Heard RH, Mellonig JT, Brunsvold MA, Lasho DJ, Meffert RM, Cochran DL. Clinical evaluation of wound healing following multiple exposures to enamel matrix protein derivative in the treatment of intrabony periodontal defects. J Periodontol 2000;71:1715-1721. Haerian A, Adonogianaki E, Mooney J, Manos A, Kinane DF. Effects of treat- ment on gingival crevicular collagenase, stromelysin and tissue inhibitor of metalloproteinases and their ability to predict response to treatment. J Clin Periodontol 1996;23:83–91. Heikkilä P, Suojanen J, Pirilä E et al. Human tongue carcinoma growth is inhibited by selective antigelatinolytic peptides. Int J Cancer 2006;118:2202-2209. Heng NH, N'Guessan PD, Kleber BM, Bernimoulin JP, Pischon N. Enamel matrix derivative induces connective tissue growth factor expression in human osteoblastic cells. J Periodontol 2007;78:2369-2379. van Hinsbergh VW, Koolwijk P. Endothelial sprouting and angiogenesis: matrix metalloproteinases in the lead. Cardiovasc Res 2008;78:203-212. Hoang AM, Klebe RJ, Steffensen B, Ryu OH, Simmer JP, Cochran DL. Amelogenin is a cell adhesion protein. J Dent Res 2002;81:497-500. Hoang AM, Oates TW, Cochran DL. In vitro wound healing responses to enamel matrix derivative. J Periodontol 2000;71:1270-1277. Hong SD, Hong SP, Lee JI, Lim CY. Expression of metalloproteinase-2 and –9 in oral squamous cell carcinomas with regard to metastatic potential. Oral Oncol 2000;36:207-213. Huang DW, Sherman BT, Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc 2009;4:44-57. Hägewald S, Pischon N, Jawor P, Bernimoulin JP, Zimmermann B. Effects of enamel matrix derivative on proliferation and differentiation of primary osteoblasts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2004;98:243-249. Impola U, Uitto VJ, Hietanen J et al. Differential expression of matrilysin-1 (MMP-7), 92 kD gelatinase (MMP-9), and metalloelastase (MMP-12) in oral verrucous and squamous cell cancer. J Pathol 2004;202:14-22. Inoue M, LeGeros RZ, Hoffman C, Diamond K, Rosenberg PA, Craig RG. Effect of enamel matrix proteins on the phenotype expression of periodontal ligament cells cultured on dental materials. J Biomed Mater Res A 2004;69:172-179. Itoh N, Kasai H, Ariyoshi W, Harada E, Yokota M, Nishihara T. Mechanisms involved in the enhancement of osteoclast formation by enamel matrix derivative. J Periodontal Res 2006;41:273-279. Iwata T, Morotome Y, Tanabe T, Fukae M, Ishikawa I, Oida S. Noggin blocks osteoinductive activity of porcine enamel extracts. J Dent Res 2002;81:387-391.

62

Jenkins G. The role of proteases in transforming growth factor-beta activation. Int J Biochem Cell Biol 2008;40:1068-1078. Jiang J, Fouad AF, Safavi KE, Spångberg LS, Zhu Q. Effects of enamel matrix derivative on gene expression of primary osteoblasts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2001;91:95-100. Jiang J, Goodarzi G, He J et al. Emdogain-gel stimulates proliferation of odontoblasts and osteoblasts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006;102:698-702. Jiang J, Safavi KE, Spangberg LS, Zhu Q. Enamel matrix derivative prolongs primary osteoblast growth. J Endod 2001;27:110-112. Johansson N, Ala-aho R, Uitto VJ et al. Expression of collagenase-3 (MMP-13) and collagenase-1 (MMP-1) by transformed keratinocytes is dependent on the activity of p38 mitogen-activated protein kinase. J Cell Sci 2000;113:227-235. Kasaj A, Willershausen B, Reichert C, Röhrig B, Smeets R, Schmidt M. Ability of nanocrystalline hydroxyapatite paste to promote human periodontal ligament cell proliferation. J Oral Sci 2008;50,279-285. Kawase T, Okuda K, Momose M, Kato Y, Yoshie H, Burns DM. Enamel matrix derivative (EMDOGAIN) rapidly stimulates phosphorylation of the MAP kinase family and nuclear accumulation of smad2 in both oral epithelial and fibroblastic human cells. J Periodontal Res 2001;36:367-376. Kawase T, Okuda K, Yoshie H, Burns DM. Cytostatic action of enamel matrix derivative (EMDOGAIN) on human oral squamous cell carcinoma-derived SCC25 epithelial cells. J Periodontal Res 2000;35:291-300. Kawase T, Okuda K, Yoshie H, Burns DM. Anti-TGF-beta antibody blocks enamel matrix derivative-induced upregulation of p21WAF1/cip1 and prevents its inhibition of human oral epithelial cell proliferation. J Periodontal Res 2002;37:255-262. Keila S, Nemcovsky CE, Moses O, Artzi Z, Weinreb M. In vitro effects of enamel matrix proteins on rat bone marrow cells and gingival fibroblasts. J Dent Res 2004;83:134-138. van Kempen LC, Rhee JS, Dehne K, Lee J, Edwards DR, Coussens LM. Epithelial carcinogenesis: dynamic interplay between neoplastic cells and their microenvironment. Differentiation 2002;70:610-623. Kerkela E, Ala-aho R, Lohi J, Grenman R, V MK, Saarialho-Kere U. Differential patterns of stromelysin-2 (MMP-10) and MT1-MMP (MMP-14) expression in epithelial skin cancers. Br J Cancer 2001;84:659–669. Kinane DF, Darby IB, Said S et al. Changes in gingival crevicular fluid matrix metalloproteinase-8 levels during periodontal treatment and maintenance. J Periodontal Res 2003;38:400-404. Klein MO, Reichert C, Koch D, Horn S, Al-Nawas B. In vitro assessment of motility and proliferation of human osteogenic cells on different isolated extracellular matrix components compared with enamel matrix derivative by continuous single-cell observation. Clin Oral Implants Res 2007;18:40-45. Knäuper V, Murphy G, Tschesche H. Activation of human neutrophil procollagenase by stromelysin 2. Eur J Biochem 1996;235:187-191. Koivunen E, Arap W, Valtanen H et al. Tumour targeting with a selective gelatinase inhibitor. Nature Biotechnology 1999;17:768-774. Krampert M, Bloch W, Sasaki T et al. Activities of the matrix metalloproteinase stromelysin-2 (MMP-10) in matrix degradation and keratinocyte organization in wounded skin. Mol Biol Cell 2004;15:5242-5254.

63

Kähäri VM, Saarialho-Kere U. Matrix metalloproteinases and their inhibitors in tumour growth and invasion. Ann Med 1999;31:34-45. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680-685. Lamster IB, Ahlo JK. Analysis of gingival crevicular fluid as applied to the diagnosis of oral and systemic diseases. Ann N Y Acad Sci 2007;1098:216-229. Lee ML, Whitmore GA. Power and sample size for DNA microarray studies. Stat Med 2002;21:3543-3570. Lijnen HR. Elements of the fibrinolytic system. Ann N Y Acad Sci 2001;936:226-236. Liotta LA, Kohn EC. The microenvironment of the tumour-host interface. Nature 2001;411:375-379. Lyngstadaas SP, Lundberg E, Ekdahl H, Andersson C, Gestrelius S. Autocrine growth factors in human periodontal ligament cells cultured on enamel matrix derivative. J Clin Periodontol 2001;28:181-188. Lyons JG, Birkedal-Hansen B, Moore WG, O'Grady RL, Birkedal-Hansen H. Characteristics of a 95-kDa matrix metalloproteinase produced by mammary carcinoma cells. Biochemistry 1991;30:1449-1456. Lόpez-Otίn C, Matrisian LM. Emerging roles of proteases in tumor suppression. Nature Rev Cancer 2007;7:800–808. Lossdörfer S, Sun M, Götz W, Dard M, Jäger A. Enamel matrix derivative promotes human periodontal ligament cell differentiation and osteoprotegerin production in vitro. J Dent Res 2007;86:980-985. Madri JA, Pratt BM, Tucker AM. Phenotypic modulation of endothelial cells by transforming growth factor-beta depends upon the composition and organization of the extracellular matrix. J Cell Biol 1988;106:1375-1384. Marcaccini AM, Meschiari CA, Zuardi LR, et al. Gingival crevicular fluid levels of MMP-8, MMP-9, TIMP-2 and MPO decreases after periodontal therapy. J Clin Periodontol 2009. Mathew R, Khanna R, Kumar R, Mathur M, Shukla NK, Ralhan R. Stromelysin-2 overexpression in human esophageal squamous cell carcinoma: potential clinical implications. Cancer Detect Prev 2002;26:222-228. Matsuda N, Horikawa M, Watanabe M, Kitagawa S, Kudo Y, Takata T. Possible involvement of extracellular signal-regulated kinases 1/2 in mitogenic response of periodontal ligament cells to enamel matrix derivative. Eur J Oral Sci 2002;110:439-444. Matsumoto K, Horikoshi M, Rikimaru K, Enomoto S. A study of an in vitro model for invasion of oral squamous cell carcinoma. J Oral Pathol Med 1989;18:498-501. Maycock J, Wood SR, Brookes SJ, Shore RC, Robinson C, Kirkham J. Characterization of a porcine amelogenin preparation, EMDOGAIN, a biological treatment for periodontal disease. Connect Tissue Res 2002;43:472-476. McCawley LJ, Matrisian LM. Matrix metalloproteinases: multifunctional contributors to tumor progression. Mol Med Today 2000;6:149-156. McCawley LJ, Matrisian LM. Matrix metalloproteinases: they're not just for matrix anymore! Curr Opin Cell Biol 2001;13:534-540.

64

McQuibban GA, Gong JH, Wong JP, Wallace JL, Clark- Lewis I, Overall CM. Matrix metalloproteinase processing of monocyte chemoattractant proteins generates CC chemokine receptor antagonists with anti-inflammatory properties in vivo. Blood 2002;100:1160-1167. Mirastschijski U, Konrad D, Lundberg E, Lyngstadaas SP, Jorgensen LN, Agren MS. Effects of a topical enamel matrix derivative on skin wound healing. Wound Repair Regen 2004;12:100-108. Mizutani S, Tsuboi T, Tazoe M, Koshihara Y, Goto S, Togari A. Involvement of FGF-2 in the action of Emdogain on normal human osteoblastic activity. Oral Dis 2003;9:210-217. Moilanen M, Sorsa T, Stenman M et al. Tumor-associated trypsinogen-2 (trypsinogen-2) activates procollagenases (MMP-1, -8, -13) and stromelysin-1 (MMP-3) and degrades type I collagen. Biochemistry 2003;42:5414-5420. Mott JD, Werb Z. Regulation of matrix biology by matrix metalloproteinases. Curr Opinn Cell Biol 2004;16:558–564. Mueller HP. Periodontology: the Essentials. Stuttgart: Thieme, 2005. Murphy G, Nagase H. Progress in matrix metalloproteinase research. Mol Aspects Med 2008;29:290-308. Mäkelä M, Larjava H, Pirilä E et al. Matrix metalloproteinase 2 (gelatinase A) is related to migration of keratinocytes. Exp Cell Res 1999;251:67-78. Mäntylä P, Stenman M, Kinane D et al. Gingival crevicular fluid collagenase-2 (MMP-8) test stick for chair-side monitoring of periodontitis. J Periodontal Res 2003;38:436-439. Mäntylä P, Stenman M, Kinane D et al. Monitoring periodontal disease status in smokers and nonsmokers using a gingival crevicular fluid matrix metalloproteinase-8-specific chair-side test. J Periodontal Res 2006;41:503-512. Nagano T, Iwata T, Ogata Y et al. Effect of heat treatment on bioactivities of enamel matrix derivatives in human periodontal ligament (HPDL) cells. J Periodontal Res 2004;39:249-256. Nagase H. Activation mechanisms of matrix metalloproteinases. Biol Chem 1997;378:151-160. Nagase H, Woessner JF. Matrix metalloproteinases. J Biol Chem 1999;274:21491–21494. Nagase H, Visse R, Murphy G. Structure and function of matrix metalloproteisases and TIMPs. Cardiovasc Res 2006;69:562–573. Nakamura H, Fujii Y, Ohuchi E, Yamamoto E, Okada Y. Activation of the precursor of human stromelysin 2 and its interactions with other matrix metalloproteinases. Eur J Biochem 1998;253:67-75. Narukawa M, Suzuki N, Takayama T, Shoji T, Otsuka K, Ito K. Enamel matrix derivative stimulates chondrogenic differentiation of ATDC5 cells. J Periodontal Res 2007;42:131-137. Nicholson R, Murphy G, Breathnach R. Human and rat malignant-tumor-associated mRNAs encode stromelysin-like metalloproteinases. Biochemistry 1989;28:5195-5203. Nyberg P, Heikkilä P, Sorsa T, et al. Endostatin inhibits human tongue carcinoma cell invasion and intravasation and blocks the activation of matrix metalloproteinases. J Biol Chem 2003;278:22404–22411. Nyberg P, Salo T, Kalluri R. Tumor microenvironment and angiogenesis. Front Biosci 2008;13:6537-6553.

65

Nyman S, Gottlow J, Karring T, Lindhe J. The regenerative potential of the periodontal ligament. An experimental study in the monkey. J Clin Periodontol 1982;9:257-265. Ohyama M, Suzuki N, Yamaguchi Y, Maeno M, Otsuka KI. Effect of enamel matrix derivative on the differentiation of C2C12 cells. J Periodontol 2002;73:543-550. Oikarinen K, Seppa S. Effect of preservation media on proliferation and collagen biosynthesis of periodontal ligament fibroblasts. Endod Dent Traumatol 1987;3:95-99. Okubo K, Kobayashi M, Takiguchi T et al. Participation of endogenous IGF-I and TGF-beta 1 with enamel matrix derivative-stimulated cell growth in human periodontal ligament cells. J Periodontal Res 2003;38:1-9. Oringer RJ. Biological mediators for periodontal and bone regeneration. Compend Contin Educ Dent 2002;23:501-512. Otsuka T, Kasai H, Yamaguchi K, Nishihara T. Enamel matrix derivative promotes osteoclast cell formation by RANKL production in mouse marrow cultures. J Dent 2005;33:749-755. Page-McCaw A, Ewald AJ, Werb Z. Matrix metalloproteinases and the regulation of tissue remodelling. Nat Rev Mol Cell Biol 2007;8:221-233. Palioto DB, Coletta RD, Graner E, Joly JC, de Lima AF. The influence of enamel matrix derivative associated with insulin-like growth factor-I on periodontal ligament fibroblasts. J Periodontol 2004;75:498-504. Pasini FS, Brentani MM, Kowalski LP, Federico MH. Transforming growth factor beta1, urokinase-type plasminogen activator and plasminogen activator inhibitor-1 mRNA expression in head and neck squamous carcinoma and normal adjacent mucosa. Head Neck 2001;23:725-732. Parkar MH, Tonetti M. Gene expression profiles of periodontal ligament cells treated with enamel matrix proteins in vitro: analysis using cDNA arrays. J Periodontol 2004;75:1539-1546. Pertovaara L, Kaipainen A, Mustonen T et al. Vascular endothelial growth factor is induced in response to transforming growth factor-beta in fibroblastic and epithelial cells. J Biol Chem. 1994;269:6271-6274. Pitaru S, McCulloch CA, Narayanan SA. Cellular origins and differentiation control mechanisms during periodontal development and wound healing. J Periodontal Res 1994;29:81-94. Pries R, Wollenberg B. Cytokines in head and neck cancer. Cytokine Growth Factor Rev 2006;17:141-146. Prime SS, Davies M, Pring M, Paterson IC. The role of TGF-beta in epithelial malignancy and its relevance to the pathogenesis of oral cancer (part II). Crit Rev Oral Biol Med 2004;15:337-347. Pääkkönen V, Vuoristo JT, Salo T, Tjäderhane L. Comparative gene expression profile analysis between native human odontoblasts and pulp tissue. Int Endod J 2008;41:117-127. Ravanti L, Kähäri VM. Matrix metalloproteinases in wound repair (review). Int J Mol Med 2000;6:391-407. Rechardt O, Elomaa O, Vaalamo M et al. Stromelysin-2 is upregulated during normal wound repair and is induced by cytokines. J Invest Dermatol 2000;115:778-787. Rincon JC, Haase HR, Bartold PM. Effect of Emdogain® on human periodontal fibroblasts in an in vitro wound-healing model. J Periodont Res 2003;38:290-295.

66

Rincon JC, Xiao Y, Young WG, Bartold PM. Enhanced proliferation, attachment and osteopontin expression by porcine periodontal cells exposed to Emdogain. Arch Oral Biol 2005;50:1047-1054. Ripamonti U, Reddi AH. Tissue engineering, morphogenesis, and regeneration of the periodontal tissues by bone morphogenetic proteins. Crit Rev Oral Biol Med 1997;8:154-163. Rodrigues TL, Marchesan JT, Coletta RD et al. Effects of enamel matrix derivative and transforming growth factor-beta1 on human periodontal ligament fibroblasts. J Clin Periodontol 2007;34:514-522. Rosenthal E, McCrory A, Talbert M, Young G, Murphy-Ullrich J, Gladson C. Elevated expression of TGF-beta1 in head and neck cancer-associated fibroblasts. Mol Carcinog 2004;40:116-121. Ruokolainen H, Pääkkö P, Turpeenniemi-Hujanen T. Expression of matrix metalloproteinase-9 in head and neck squamous cell carcinoma: a potential marker for prognosis. Clin Cancer Res 2004;10:3110-3116. Ruokolainen H, Pääkkö P, Turpeenniemi-Hujanen T. Tissue and circulating immunoreactive protein for MMP-2 and TIMP-2 in head and neck squamous cell carcinoma--tissue immunoreactivity predicts aggressive clinical course. Mod Pathol 2006;19:208-217. Rusciano D; Burger M. Why do cancer cells metastasize into particular organs? BioEssays 1992;14:185-194. Ryu OH, Fincham AG, Hu CC, et al. Characterization of recombinant pig enamelysin activity and cleavage of recombinant pig and mouse amelogenins. J Dent Res 1999;78:743-750. Saari H, Suomalainen K, Lindy O, Konttinen YT, Sorsa T. Activation of latent human neutrophil collagenase by reactive oxygen species and serine proteases. Biochem Biophys Res Commun 1990;171:979-987. Saarialho-Kere UK. Patterns of matrix metalloproteinase and TIMP expression in chronic ulcers. Arch Dermatol Res 1998;290:S47-54. Salo T, Liotta L, Keski-Oja J, Turpeenniemi-Hujanen T, Tryggvason K. Secretion of basement membrane collagen degrading enzyme and plasminogen activator by transformed cells- role in metastasis. Int J Cancer 1982;30:669-673. Salo T, Liotta L, Tryggvason K. Purification and characterization of a murine basement membrane collagen-decrading enzyme secreted by metastatic tumor cells. J Biol Chem 1983; 258: 3058–3063. Salo T, Lyons JG, Rahemtulla F, Birkedal-Hansen H, Larjava H. Transforming growth factor-beta 1 up-regulates type IV collagenase expression in cultured human keratinocytes. J Biol Chem. 1991;266:11436-11441. Salo T, Mäkelä M, Kylmäniemi M, Autio-Harmainen H, Larjava H. Expression of matrix metalloproteinase-2 and -9 during early human wound healing. Lab Invest 1994;70:176-182. Sauter ER, Nesbit M, Watson JC, Klein-Szanto A, Litwin S, Herlyn M. Vascular endothelial growth factor is a marker of tumor invasion and metastasis in squamous cell carcinomas of the head and neck. Clin Cancer Res 1999;5:775-782. Schlueter SR, Carnes DL, Cochran DL. In vitro effects of enamel matrix derivative on microvascular cells. J Periodontol 2007;78:141-151. Schwartz Z, Carnes DL Jr, Pulliam R et al. Porcine fetal enamel matrix derivative stimulates proliferation but not differentiation of pre-osteoblastic 2T9 cells, inhibits proliferation and stimulates differentiation of osteoblast-like MG63 cells, and increases proliferation and differentiation of normal human osteoblast NHOst cells. J Periodontol 2000;71:1287-1296.

67

Schwarz F, Rothamel D, Herten M, Sculean A, Scherbaum W, Becker J. Effect of enamel matrix protein derivative on the attachment, proliferation, and viability of human SaOs(2) osteoblasts on titanium implants. Clin Oral Investig 2004;8:165-171. Schwingshackl A, Duszyk M, Brown N, Moqbel R. Human eosinophils release matrix metalloproteinase-9 on stimulation with TNF-alpha. Allergy Clin Immunol 1999;104:983-989. Seargent JM, Loadman PM, Martin SW, Naylor B, Bibby MC, Gill JH. Expression of matrix metalloproteinase-10 in human bladder transitional cell carcinoma. Urology 2005;65:815-820. Sheu BC, Lin RH, Lien HC, Ho HN, Hsu SM, Huang SC. Predominant Th2/Tc2 polarity of tumor-infiltrating lymphocytes in human cervical cancer. J Immunol 2001;167:2972-2978. Shimizu E, Nakajima Y, Kato N et al. Regulation of rat bone sialoprotein gene transcription by enamel matrix derivative. J Periodontol 2004;75:260-267. Shipley JM, Doyle GA, Fliszer CJ et al. The structural basis for the elastolytic activity of the 92-kDa and 72-kDa gelatinases. Role of the fibronectin type II-like repeats. J Biol Chem 1996;271:4335-4341. Sinpitaksakul SN, Pimkhaokham A, Sanchavanakit N, Pavasant P. TGF-beta1 induced MMP-9 expression in HNSCC cell lines via Smad/MLCK pathway. Biochem Biophys Res Commun 2008;371:713-718. Slavkin HC. Positional signalling and patterning for amelogenesis in mouse molar tooth development. Connect Tissue Res 1989;20:91-100. Smith B, Smith G, Carter D, Sasaki C, Haffty BG. Prognostic significance of vascular endothelial growth factor protein levels in oral and oropharyngeal squamous cell carcinoma. J Clin Oncol 2000;18:2046–2052. Smith J, Rattay T, McConkey C, Helliwell T, Mehanna H. Biomarkers in dysplasia of the oral cavity: a systematic review. Oral Oncol 2009;45:647-653. Sorsa T, Ingman T, Suomalainen K et al. Identification of proteases from periodontopathogenic bacteria as activators of latent human neutrophil and fibroblast-type interstitial collagenases. Infect Immun 1992;60:4491-4495. Sorsa T, Tjäderhane L, Konttinen YT, et al. Matrix metalloproteinases: contribution to pathogenesis, diagnosis and treatment of periodontal inflammation. Ann Med 2006;38:306-321. Sorsa T, Tjäderhane L, Salo T. Matrix metalloproteinases (MMPs) in Oral Diseases. Oral Dis 2004;10:311-318. Song AM, Shu R, Xie YF et al. A study of enamel matrix proteins on differentiation of porcine bone marrow stromal cells into cementoblasts. Cell Prolif 2007;40:381-396. Spahr A, Hammarström L. Response of dental follicular cells to the exposure of denuded enamel matrix in rat molars. Eur J Oral Sci 1999;107:360-367. Sterne J, Davey Smith G. Shifting the evidence – what’s wrong with significance tests. BMJ 2001;322:226-231. Sternlicht MD, Lochter A, Sympson CJ et al. The stromal proteinase MMP3/stromelysin-1 promotes mammary carcinogenesis. Cell 1999;87:137-146.

68

Sternlicht MD, Werb Z. How matrix metalloproteinases regulate cell behaviour. Annu Rev Cell Biol 2001;17:463–516. Stetler-Stevenson WG, Yu AE. Proteases in invasion: matrix metalloproteinases. Cancer Biol 2001;11:143–152. Stöcker W, Bode W. Structural features of a superfamily of zinc-endopeptidases: the metzincins. Curr Opin Struct Biol 1995;5:383-390. Suzuki S, Nagano T, Yamakoshi Y, et al. Enamel matrix derivative gel stimulates signal transduction of BMP and TGF-{beta}. J Dent Res 2005;84:510-514. Sutinen M, Kainulainen T, Hurskainen T et al. Expression of matrix metalloproteinases (MMP-1 and -2) and their inhibitors (TIMP-1, -2 and -3) in oral lichen planus, dysplasia, squamous cell carcinoma and lymph node metastasis. Br J Cancer 1998;77:2239-2245. Takayama T, Suzuki N, Narukawa M, Goldberg HA, Otsuka K, Ito K. Enamel matrix derivative is a potent inhibitor of breast cancer cell attachment to bone. Life Sci 2005;76:1211-1221. Takayanagi K, Osawa G, Nakaya H, Cochran DL, Kamoi K, Oates TW. Effects of enamel matrix derivative on bone-related mRNA expression in human periodontal ligament cells in vitro. J Periodontol 2006;77:891-898. Thomas GT, Lewis MP, Speight PM. Matrix metalloproteinases and oral cancer. Oral Oncol 1999;35:227-233. Tokiyasu Y, Takata T, Saygin E, Somerman M. Enamel factors regulate expression of genes associated with cementoblasts. J Periodontol 2000;71:1829-1839. Van den Steen PE, Dubois B, Nelissen I, Rudd PM, Dwek RA, Opdenakker G. Biochemistry and molecular biology of gelatinase B or matrix metalloproteinase-9 (MMP-9). Crit Rev Biochem Mol Biol 2002;37:375-536. Vartio T, Hovi T, Vaheri A. Human macrophages synthesize and secrete a major 95,000-dalton gelatine-binding protein distinct from fi bronectin. J Biol Chem 1982;257:8862–8866. de Vicente JC, Fresno MF, Villalain L, Vega JA, Hernandez Vallejo G. Expression and clinical significance of matrix metalloproteinase-2 and matrix metalloproteinase-9 in oral squamous cell carcinoma. Oral Oncol 2005;41:283-293. Vihinen P, Kahari VM. Matrix metalloproteinases in cancer: prognostic markers and therapeutic targets. Int J Cancer 2002;99:157–166. Visse R, Nagase H. Matrix metalloproteinases and tissue inhibitors of metalloproteinases: structure, function, and biochemistry. Circ Res 2003;92:827–839. Vu TH, Werb Z. Matrix metalloproteinases: effectors of development and normal physiology. Genes Dev 2000;14:2123-2133. Westermarck J, Kähäri VM. Regulation of matrix metalloproteinase expression in tumor invasion. FASEB J 1999;13:781-792. Wielockx B, Lannoy K, Shapiro SD et al. Inhibition of matrix metalloproteinases blocks lethal hepatitis and apoptosis induced by tumor necrosis factor and allows safe antitumor therapy. Nat Med 2001;7:1202-1208.

69

Wilkins-Port CE, Higgins PJ. Regulation of extracellular matrix remodeling following transforming growth factor-beta1/epidermal growth factor-stimulated epithelial-mesenchymal transition in human premalignant keratinocytes. Cells Tissues Organs 2007;185:116-22. Windsor LJ, Grenett H, Birkedal-Hansen B, Bodden MK, Engler JA, Birkedal-Hansen H. Cell type-specific regulation of SL-1 and SL-2 genes. Induction of the SL-2 gene but not the SL-1 gene by human keratinocytes in response to cytokines and phorbolesters. J Biol Chem 1993;268:17341-17347. Wyke JA. Overview – burgeoning promise in metastasis research. Eur J Cancer 2000;36:1589-1594. Yu Q, Stamenkovic I. Cell surface-localized matrix metalloproteinase-9 proteolytically activates TGF-beta and promotes tumor invasion and angiogenesis. Genes Dev 2000;14:163-176. Yuan K, Chen CL, Lin MT. Enamel matrix derivative exhibits angiogenic effect in vitro and in a murine model. J Clin Periodontol 2003;30:732-738. Zeichner-David M. Regeneration of periodontal tissues: cementogenesis revisited. Periodontol 2000 2006;41:196-217. Zeichner-David M. Is there more to enamel matrix proteins than biomineralization? Matrix Biol 2001;20:307-316. Zeichner-David M, Chen LS, Hsu Z, Reyna J, Caton J, Bringas P. Amelogenin and ameloblastin show growth-factor like activity in periodontal ligament cells. Eur J Oral Sci 2006;114:244-253. Zeldich E, Koren R, Nemcovsky C, Weinreb M. Enamel matrix derivative stimulates human gingival fibroblast proliferation via ERK. J Dent Res 2007;86:41-46. Zetterström O, Andersson C, Eriksson L et al. Clinical safety of enamel matrix derivative (EMDOGAIN) in the treatment of periodontal defects. J Clin Periodontol 1997;24:697-704.

70

APPENDIX The 15 most enriched GO terms according to P-value by 6 h (A1/A2) of EMD and 6 h (B1/B2) and 24 h (C1/C2) of TGF- β1 treatments of HSC-3 cells. Table shows the term, p-value and number of genes (N. G.) related to various terms. * correspond to terms, which were the same for EMD and TGF-β1 after 6 h of treatment. “ correspond to terms, which were the same for 6 h and 24 h of TGF-β1 treatments.