UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Experimental diabetes in ß6 integrin deficient mice Aurora, Saljae 2009

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2009_fall_aurora_saljae.pdf [ 2.84MB ]
JSON: 24-1.0068196.json
JSON-LD: 24-1.0068196-ld.json
RDF/XML (Pretty): 24-1.0068196-rdf.xml
RDF/JSON: 24-1.0068196-rdf.json
Turtle: 24-1.0068196-turtle.txt
N-Triples: 24-1.0068196-rdf-ntriples.txt
Original Record: 24-1.0068196-source.json
Full Text

Full Text

EXPERIMENTAL DIABETES IN 136 INTEGRIN DEFICIENT MICE by SALJAE AURORA B.Sc., The University of Alberta, 1988 D.D.S., The University of Alberta, 1992 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Dental Science) THE UNIVERISTY OF BRITISH COLUMBIA (Vancouver) July 2009 © Saljae Aurora, 2009 ABSTRACT Objectives: Periodontal disease involves transformation of the junctional epithelium (JE) to a pocket epithelium (PE). Integrin av136 is constitutively expressed in the healthy IF but not in PE. Mice deficient of this integrin (136-I-) exhibit increased periodontal bone loss and PE migration. Thus, ctvf36 integrin in JE may have a protective role against periodontal disease. As diabetes aggravates periodontal disease, we hypothesized that diabetic 136-I- mice would develop more advanced periodontal disease compared to non- diabetic mice. Furthermore, we hypothesize that re-introduction of 136 integrin under the cytokeratin 14 (Ki 4) gene promoter in the 136-I- strain, would restore protection against periodontal disease. Methods: Wild-type (WIT), f36-I-, 136 integrin over-expressing (Kl4136; f36 integrin expression is driven by Ki 4 promoter) and f36 integrin rescue (f36-rescue; cross-breed of f36-/- and K14f36 mice) mice were induced to develop diabetes by injections of streptozotocin, and confirmed diabetic 2 weeks later (Blood glucose >306 mg/dl (17 mmollL). Control animals were exposed to the citrate vehicle only. Four months later, mice were sacrificed. Mandibles were dc-fleshed and imaged for quantification of periodontal bone loss. Maxillae were decalcified and sectioned for histological assessment of epithelial migration and inflammation. Results: Diabetic 136-I- mice (71% mortality) exhibited significantly higher death rates as compared to non-diabetic WT and K14j36 groups. Epithelial migration for the 136-/ strain was significantly higher than that seen in the W/T and the K14136 strains. (P<0.001). Significantly reduced epithelial migration was noted for the 136 Rescue strain compared to 136-I- strain. Periodontal bone loss was significantly greater in the male 136- 11 I- strain when combined with experimental diabetes (P<O.05), while experimental diabetes did not increase alveolar bone loss in other male groups. Female diabetic mice appear protected from alveolar bone loss. Inflammation was significantly greater for both the 136 Rescue strain and 136-I- strain relative to the W/T strain. Conclusions: Integrin av136 protects diabetic animals from death and periodontal disease. Reductions in epithelial migration and mortality in the 136 Rescue strain reinforce the role of cxv136 integrin in suppressing inflammation. Diabetes can aggravate periodontal bone loss in male diabetic 136-I- mice. 111 TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables vii List of Figures ix List of Abbreviations xv Acolegeents xviii Dedication xx CHAPTER 1: INTRODUCTION 1 1.1 The Periodontium in Health and Disease 1 1.1.1 Gingival Epithelium 2 Junctional Epithelium 3 1.1.2 Gingival Connective Tissue 6 1.1.3 Pocket Epithelium 7 1.2 Inflammation in the Periodontium 9 1.2.1 Bacterial Plaque 9 1.2.2 Gingival Inflammation-Acute 9 1.2.3 Gingival Inflammation-Chronic 11 1.2.4 Mediators of Inflammation-Cytokines 12 Transforming Growth Factor 5 14 1.3 Integrin av6 17 1.3.1 Genetic Knock-out Studies 22 1.3.2 Integrin avf36 and the Junctional Epithelium 24 iv 1.4 Diabetes Mellitus and Periodontal Disease 25 1.5 Hypothesis, Aim and Objectives of the Study 30 CHAPTER 2; MATERIALS AND METHODS 32 2.1 Induction of Experimental Diabetes Mellitus 32 2.2 Tissue Preparation and Assessment 34 2.2.1 Mandible 34 Quantification of Root Exposure and Crown Width 35 2.2.2 1’Iaxil1a 37 Sectioning of Embedded Tissues 38 Quantification of Epithelial Migration 38 Assessment of Periodontal Inflammation 40 Assessment of Epithelial Migration into the Furcation 42 2.2.3 Characterization of 6 Integrin Transgenic and Gene Knock-out Mouse Phenotypes by PCR 43 2.3 Blinding of Examiner 45 2.4 Statistics 46 CHAPTER 3: IESI.JI_4TS 46 3.1 Survival Analysis 47 3.2 Pretreatment Weight 49 3.3 Experimental Diabetes 50 3.4 Epithelial Migration 51 3.5 Epithelial Migration into the Furcation 54 3.6 Alveolar Bone Loss 55 V TA 801XIfflJIJJV £6S[1H131 Z6!PS .IOJSUOiJ1pUWtUO)JJ 16soisrnio:çjjyj L8i(PIIJSJoSUO!pLU!J’ 89:j L9SHIrjSflOjAJflO -)OUNpu!USU1TJ1qjHiUO!SS3.IdX3U!.IJUI911JoU0!J1IT1138 J,9UO!WWUIJUIIUA!U!OLE £93PI4T —SSO’]MIO.IIOAIVE9 Z9 — SSO1UOil.IUI0AIVZ9E 09SSOUOflI1TWIPUtI’IATpU!qWO3J9 LIST OF TABLES Table 1. Qualitative assessment of levels of gingival leukocytic infiltrate as scored from histologic sections of the interproximal regions of the molar teeth 41 Table 2. Presentation of numbers of animals surviving the 16-week experimental period and total numbers of animals that began the experimental period 47 Table 3. Epithelial migration as measured at the Ml distal, M2 mesial and M2 distal tooth surfaces. Measurements performed on digital images taken of paraffin sections and measured using ImageJ software. Number for each group represents numbers of surfaces measured in each experimental group 52 Table 4. Proportion of sites in each experimental group, for which epithelium was noted in the Ml and M2 furcation regions as determined histologically on maxillary sections. Data is an aggregate of the Ml and M2 sites. “n” represents the total number of furcation areas evaluated in each group 55 Table 5. First molar mesio-distal tooth width as measured from de-fleshed and stained mandibular jaws digitally imaged under 40x magnification and measured using ImageJ software. Statistical analysis was performed using one-way ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.01, (***): P-value <0.001 58 vii lilA 09iooo>nTA-d:()‘too>nIA-a:()‘coo>nJA-J:() iurnsn1poqsod!UUOJJUOHVAOMVXM-UOU!Sflpuuojidsouaijjp 1pu;iojsisAjuiswsijosf1l.uIuisnpinsiwpuPUOPflUPUI xo1.ipunpPwijppSMPI.IPjnqTpuPwputrnspupsjj-pwoijpossoss ‘1sJijoojpuf’JDuooAs.qiaiipspgunbSSOJuoqIPOWSJTJ9 LIST OF FIGURES Figure 1. Experimental protocol for induction of streptozotocin induced diabetes in the experimental groups and comparable treatments in the control groups of mice 34 Figure 2. Quantification of the root surface area between the CEJ and alveolar crest of the first molar using ImageJ software (40x magnification). Jaws have been stained to facilitate identification of the CEJ and alveolar crest. Images taken with the camera lens oriented perpendicular to the root surface. All quantification was done at a constant magnification, by one examiner and reliability measurements were verified to be highly significant (Pearson correlation P-value=O.942) 36 Figure 3. First molar crown width measurement using ImageJ software (40x magnification). Jaws have been stained to facilitate identification of the CEJ and alveolar crest. Images taken with the camera lens oriented perpendicular to the root surface. All quantification was done at a constant magnification, by one examiner and reliability measurements were verified to be highly significant (Pearson correlation P-valueO.979) 37 Figure 4. Representative H&E stained section between first and second molar (lOx). CEJ (grey arrow heads). Apical extent of epithelial migration (white arrow heads). Length of epithelial migration measured for assessment (dashed line). Bar200 micrometers 40 ix Figure 5. Representative H&E stained sections (lOx magnification) of the inter-proximal gingiva and alveolar crest between the first and second molar used for qualitative scoring of gingival inflammation. Areas evaluated were immediately subjacent to the gingival epithelium and pocket epithelium. Bar= 250 micrometers. A. Score 1-Mild inflammation B. Score 2-Moderate inflammation C. Score 3-Severe inflammation 42 Figure 6. Epithelial migration into furcation region as assessed on sagital sections of maxillary molar teeth. A) H&E staining of the mid bucco-lingual section of Ml showing intact periodontal ligament and no evidence of epithelial migration. B) H&E staining of the bucco-lingual mid section of Ml with evidence of epithelial migration (arrow). Bar250jim 43 Figure 7. Graphical survival analysis for all groups during the 16-week experimental period. Statistical analysis (ANOVA with a Bonferonni post hoc adjustment) revealed a significant difference for the experimental B6-/- group from both the W/T Diabetic and W/T Control groups. (*): P-value 0.05 48 Figure 8. Assessment of pretreatment weight for different strains ± one standard deviation. Animals were weighed prior to any interventions. Statistical analysis was performed using ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.01, (***): P-value =0.001 49 x Figure 9. Fasting blood glucose of all animals during experimental period as assessed with tail blood samples ± standard deviations. Differences between groups was determined using ANOVA with a Bonferonni post hoc adjustment. No differences were noted among the different diabetic groups. Similarly, no significant differences were noted among the control groups. (*): P-value <0.05, (**): P-value<0.O1, (***): P-value <0.001, (****): P-value <0.0001 51 Figure 10. Epithelial migration as measured at the Ml distal, M2 mesial and M2 distal tooth surfaces. Measurements performed on paraffin sections measured using ImageJ software. Number for each group represents numbers of surfaces measured in each experimental group. Statistical differences between groups were identified using ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.01, (***): P-value <0.001 53 Figure 11. Epithelial migration as measured at the Ml distal, M2 mesial and M2 distal tooth surfaces ± standard deviation. Measurements performed on paraffin sections measured using ImageJ software. Number for each group represents numbers of surfaces measured. Statistical analysis doneusing ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.01, (***): P-value <0.001, (****): P-value<0.0001 54 Figure 12. Proportion of Ml and M2 furcation regions that exhibited epithelial tissues in the furcation as measured histologically in the maxilla. Numbers of molar teeth xi assessed in each group is presented. Statistical differences between groups were identified using two-tailed Fisher’s exact test. (#): P-value 0.07, (*): P-value <0.05, (**): P-value<0.Ol, (***): P-value <0.001 56 Figure 13. First molar mesio-distal tooth width as measured from de-feleshed and stained mandibular jaws digitally imaged under 40x magnification and measured using ImageJ software. Statistical differences between groups were identified using ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.01, (***): P value =0.001 59 Figure 14. Quantification of surface area between the CEJ and alveolar crest of the first molar ± standard deviation as measured by ImageJ software. Statistical analysis for differences between groups supplied by ANOVA with a Bonferonni post hoc adjustment. No differences were noted between control and experimental groups within each strain. Number for each group represents numbers of teeth examined. (*): P-value <0.05, (**): P-value <0.01, (***): P-value 0.001 61 Figure 15. Quantification of surface area between the CEJ and alveolar crest of the first molar alveolar bone loss ± standard deviation for female mice as measured from digital images by ImageJ software. Statistical analysis provided by pair-wise comparisons using Student t-test. (#): P-value 0.05-0.08, (*): P-value <0.05, (***): P-value =0.001 62 xii Figure 16. Quantification of surface area between the CEJ and alveolar crest of the first molar alveolar bone loss ± standard deviation for female mice as measured from digital images by ImageJ software. Statistical analysis provided by pair-wise comparisons using Student t-test. (*): P-value <0.05, (**): P-value <0.001, (***): P-value =0.001 64 Figure 17. Qualitative assessment of inflammation immediately subjacent to the JE and/or PE as measured histologically from from 30 animals. 30 interproximal sites evaluated at the M1/M2 region and 29 interproximal sites evaluated at the M2/M3 region. Fesults from both the M1/M2 and M2/M3 regions aggregated together. Statistical analysis performed using Mann-Whitney U test (pair-wise assessments). (**): P-value<0.01, (***): P-value =0.001 65 Figure 18. Qualitative assessment of inflammation immediately subjacent to the JE and/or PE as measured histologically from sagitally sectioned maxilla from 30 animals. Experimental and control groups for each strain were combined into mouse strain groups. (30 interproximal sites evaluated at Ml & M2, 29 interproximal sites evaluated at M2 & M3) Statistical analysis performed using Mann-Whitney U test (pair-wise assessments). No statistical difference was noted between the W/T strain and the K14B6 strain. (#): P value =0.06, (*): P-value <0.05, (**): P-value<0.01, (***): P-value<0.001 66 Figure 19. Characterization of the t36 integrin transgenic and knock-out mouse lines by PCR and RT-PCR. The PCR products were separated by electrophoresis in a 1% agarose gel, stained with ethidium bromide and visualized under a UV light. Results confirm that xlii f36-/- strain is deficient in murine and human f36 integrin genes while K14 f36 strain express both murine and human 136 integrin genes. The f36 Rescue strain expresses only the human f36 integrin gene while the wild-type mice express the mouse 136 integrin only 67 xiv LIST OF ABBREVIATIONS AGE Advanced glycation end-products (36-I- 136 integin deficient (36 Rescue Progeny of K14(36 and (36-I- CCD Charge-coupled device CEJ Cemento-enamel junction DAT Directly attached to the tooth DKA Diabetic ketoacidosis ECM Extracellular matrix EBL External basal lamina EDTA Ethylenediamine tetracetic acid ERK Extracellular signal regulated kinase GTPases Guanosine triphosphoesterases Hb Hemoglobin h(36 Human transgene for the (36 sub-unit IBL Internal basal lamina IL Interleukin INF-y Interferon-y JE Junctional epithelium iNK Jun N-terminal kinase JPEG Joint photographic experts group K 14 Cytokeratin 14 xv K14f36 Cytokeratin 14 driven f36 integrin expression LAP Latency associated peptide LLC Large latent complex LPS Lipopolysaccharide LTA Lipotechoic Acid LTBP Latent TGF3 binding protein Ml First molar MiD Distal surface of Ml M2 Second molar M2D Distal surface of M2 M2M Mesial surface of M2 M3 Third molar MCP-1 Monocyte Chemo-attractant protein-i MIVIP Matrix metalloproteinases mRNA Messenger ribonucleic acid OPG Osteoprotegerin PAMPs Pattern recognition molecular patterns PBS Phosphate Buffered Saline Pg Porphyromonas gingivalis PGE-2 Prostaglandin-E2 PGN Peptidoglycans PKC Protein Kinase C PRE Preexisting root exposure xvi PRR . Pattern recognition receptors RAGE Receptor for AGE R.ANKL Receptor activator of nuclear factor KB REE Reduced enamel epithelium RGD Arginine-glycine-aspartic acid RGE Arginine-glycine-glutamic acid ROS Reactive oxygen species SLC Small latent complex sRAGE Soluble RAGE T3R TGF3 receptors Td Treponema denticola TESPA (3 -aminopropyl) triethoxysilane Tf Tannerellaforsythia Th T-helper cells Thi T-helper 1 cells Th2 T-helper 2 cells TIFF Tagged image file format TIMPs Tissue inhibitors of metalloproteinases TGF Transforming growth factor 13 TLR Toll-like receptors TNF-a Tissue necrosis factor-a TSP-l Thrombospondin-l WIT Wild type xvii ACKNOWLEDGEMENT Though this work bears my name and I accept all responsibility for its content and accuracy, I must acknowledge the contribution of those, without whom, it would not have been possible. I humbly thank my supervisor Dr. Haimu Larjava for his thoughtful guidance in the execution of the investigations reported here and preparation of this document. Similarly, for his willingness to respond to questions and constructively respond to my thoughts, I thank Dr. Lan Häkkinen. It is difficult to encompass the extent of the contribution that Mr. Cristian Sperantia has made to this project, but his patience, selflessness and above all else, his attention to detail, have earned my respect. Dr. Carol Oakley must be recognized for selfless pursuit of excellence in clinical care, and clinical education. I feel privileged to have had her instruct and guide me over the past three years. My laboratory colleagues, Dr. Leeni Koivisto, Dr. Ameneh Eslami, Dr. Yangshuang Xie, Dr. Gethin Owen, Dr. Guoqiao Jiang, Dr. Farzin Ghannad and Mr. Andre Wong have earned my respect for sharing their thoughts and time with me. xviii Dr. Tassos Irinakis, Ms. Vicky Koulouris, Ms. Miriam Dexter, Ms. Nancy Per! and Ms. Cherry Ong have been supportive throughout the program and for that I am thankful. Dr. Christopher Chung, my classmate, has walked this path with me step by step, often quietly leading the way, and has exemplified what it means to be a friend. I thank you sincerely for everything. Most importantly, I must recognize the selflessness of my wife, Manishi, and the understanding of my daughters, Menyka and Ayshani, throughout the last 3 years. They have been my light during every waking moment, the inspiration for every written word and the strength for each stride taken towards the completion of these studies. Their smiles, hugs and kisses embody every word and thought presented here. Supported by a grant from CIHR xix DEDICATION To my parents, Om and Shanta, this body of work is dedicated to you. The immeasurable sacrifices associated with leaving your families and emigrating were made with the goal of securing higher education and better futures for your children. I will never truly know what you have given up but am grateful to you for everything, including your insight, wisdom and unwavering love and support. I aspire to embody these virtues in my years ahead and will always turn back to my experiences and upbringing for guidance while I look towards the future. xx CHAPTER 1: INTRODUCTION 1.1 The Periodontium in Health and Disease Periodontal disease represents the culmination of bacterially initiated infLammation in a susceptible host resulting in loss of periodontal attachment and deepening of the gingival sulcus (Kinane and Attstrom, 2005). More-over, many systemic diseases have been associated with periodontal disease and evidence is mounting implicating periodontal disease as having a pathophysiologic role in these diseases (Kantarci and Van Dyke, 2005). Clinically observable gingival inflammation is a predictable sequelae following accumulations and maturation of dental bacterial biofllm (Loe et al., 1965). Periodontal disease becomes clinically observed when the dentition experiences loss of attachment associated with deepening of the gingival sulcus. Histologically, periodontal attachment loss is preceded by the transformation of the clinically normal junctional epitheliuin to a pathological pocket epithelium (Muller-Glauser and Schroeder, 1982; Page and Schroeder, 1976). While bacterial plaque is necessary to initiate gingival inflammation, the presence of plaque alone seems insufficient for the predictable transformation of gingival disease to periodontal disease, implicating the necessity of some modulating host response factors (Page and Schroeder, 1976). The junctional epithelium is critically situated at the juncture of the bacterial plaque and the host defenses. Under low-grade plaque induced inflammatory pressures it is relatively stable but at certain locations and/or individuals, junctional epithelium loses its structure to become a migratory pocket epithelium, a change that is critical for the progression of gingival disease to periodontal disease. Its stability, in the face of chronic inflammation, is central to management of 1 this bacterial threat and variations in this ability to withstand the inflammatory stress may represent a risk factor for the progression of periodontal disease. Understanding the factors associated with susceptibility to periodontal attachment loss is necessary before developing strategies for effective treatment and prevention of periodontal disease and ultimately reducing overall risk from the co-morbidities from the associated systemic diseases. 1.1.1 Gingival Epithelium Fully erupted healthy teeth are clinically surrounded by gingival epithelium attached to the tooth immediately coronal to the cemento-enamel junction (CEJ). This gingiva is broadly classified into three phenotypes (Bartold et al., 2000). The oral gingival epithelium is a keratinized tissue that extends from the muco-gingival junction to the gingival crest. The sulcular epithelium is non-keratinized under normal (non experimental) conditions and extends from the gingival crest to the base of the gingival sulcus. The junctional epithelium (JE), also a non-keratinized epithelium, attaches to the tooth surface at the base of the sulcus. Unlike the oral gingival epithelium which expresses well developed rete ridges, the junctional epithelium and oral sulcular epithelium are generally devoid of well-developed rete ridges with a smoother undulating interface with the connective tissue (Bosshardt and Lang, 2005; Fiorellini, 2007; Schroeder and Listgarten, 1997). Histologically, these epithelial phenotypes seamlessly blend from one to another but they are structurally quite distinct, an observation that is apparent by their cytokeratin expression profile. 2 Cytokeratins are a group of intermediate filaments contributing to the cellular cytoskeleton of the epithelial cells (Omary et al., 2004). The 20 different cytokeratin proteins, divided into acidic and neutral/basic categories, form heterodimers and generally create an interconnecting network that extends from the plasma membrane surface to the cells nuclear surface (Chu and Weiss, 2002; Hunter et al., 2001). Cytokeratins primarily function to provide cellular protection from mechanical stress and apoptosis, though emerging evidence suggests that cytokeratins may have roles in cellular signaling, availability of cellular proteins and targeting of cellular proteins in polarized cells (Coulombe and Omary, 2002). In this manner, cytokeratins are useful in identifying cellular phenotype in continuous epithelium, level and type of differentiation and may imply functional state (Bampton et al., 1991; Jiang and Li, 2005). Cytokeratin 14 is an acidic type 1 keratin that in health is identifiable in the junctional epithelium and basal cells of oral sulcular epithelium, but when associated with advanced periodontitis shows expression at all epithelials layer and cells including the pocket epithelium (Hunter et al., 2001; Pritlove-Carson et al., 1997). Junctional Epithelium The mature junctional epithelium is comprised of stratified squamous non-keratinizing epithelium (Schroeder and Listgarten, 2003). In healthy human samples, the junctional epithelium has been measured to be 0.97mm with a range of 0.71 mm to 1.35 mm (Garguilo, 1961; Ingber et al., 1977). The junctional epithelium tapers from its coronal aspect, where it is 15-30 cells thick, to its apical extent, where it is 1-3 cells thick, about 2mm from the alveolar bone crest (Bartold et al., 2000). The junctional epithelium is 3 composed of two cell types, the basal cells (stratum basale) which face the gingival connective tissue, and the suprabasal layers (stratum suprabasale) which face the tooth (Bosshardt and Lang, 2005). The junctional epithelium abuts the subjacent connective tissue via the external basal lamina (EBL) and the tooth by the internal basal lamina (IBL) (Oksonen et al., 2001). The IBL is considered a classic basement membrane as it lacks many of the classic extracellular basement membrane components (collagen IV and laminin-1) and instead is composed almost exclusively of laminin-5 and some type VIII collagen (Oksonen et al., 2001; Pollanen et al., 2003). The suprabasal cells facing the internal basal lamina which are directly attached to the tooth (DAT), are responsible for the production of this pseudo-basement membrane structure and share some similarities with basal cells including the ability to form an attachment and representing a distinct population of proliferating cells (Kobayashi et al., 1976; Overman and Salonen, 1994; Pbllanen et al., 2003). The DAT cells are connected to the IBL via hemidesmosomes (Hormia et al., 2001; Shimono et al., 2003). Cells of the junctional epithelium undergo rapid turnover and renewal (Pollanen et a!., 2003). This turnover of cells has been shown to result in a general flow of cells towards the gingival sulcus where they exfoliate (Pollanen et a!., 2003; Willberg et a!., 2006). The rate of turnover has been estimated to be 50-100 times faster than the oral epithelial cell renewal rate, taking 4-6 days in monkeys (Shimono et al., 2003). The rate and direction conceivably help stem the tide against bacterial penetration, and the replenishment aids in turning over cells maintaining tissue homeostasis (Pollanen et al., 2003). 4 The junctional epithelium is derived primarily from the reduced enamel epithelium (REE) (Bosshardt and Lang, 2005). As the erupting tooth and follicle move closer to the mucosal surface, the papillary cells adjacent to the REE undergo a morphogenesis becoming similar to the epithelial cells (Hunter et al., 2001). Interstitial cells between the REE and the oral mucosal epithelium undergo apoptosis and the papillary cells fuse with the oral epithelium to become the primary junctional epithelium (Shimono et al., 2003). If the junctional epithelium is surgically removed, it is regenerated within a couple of weeks, and the cells that contribute to its reformation are largely derived from the oral gingival epithelium with some contribution possibly from the epithelial rests of Malassez (Locke et al., 2008; Waerhaug, 1978). In contrast to the tight interface with the tooth via the internal basal lamina and the tight intercellular junctions of the adjacent sulcular epithelium, the ultra-structural nature of the cells of the junctional epithelium suggests a looser intercellular adhesive relationship. The intercellular spaces in the junctional epithelium are almost 100% to 150% greater than the inter-cellular spaces that exist in other oral mucosal epithelium (Shimono et al., 2003). The increased inter-cellular space in the junctional epithelium is complemented by fewer and smaller desmosomes (Hatakeyama et al., 2006). The junctional epithelial spaces are frequently associated with migrating neutrophils, and the extra-cellular spaces facilitate easy passage. 5 1.1.2 Gingival Connective Tissue The gingival epithelium is supported by the connective tissue of the lamina propria, from which it is separated by the basement lamina. Cellular composition of the lamina propria is primarily fibroblasts, often spindle shaped cells, individually interspersed throughout an extra-cellular matrix (ECM) of primarily collagens, mostly type I, and proteoglycans (Bartold et al., 2000; Locke et al., 2008). Blood vessels are a prominent feature of the gingival connective tissue and the gingival tissues represent one of the largest end-organ blood supplies in the body (Bartold et al., 2000). Of the two distinct networks associated with the gingiva, one network, the gingival plexus, is located adjacent to the junctional epithelium interface, facilitating a reliable proximal supply of immune cells and serum to sustain the chronic inflammatory lesion around the junctional epithelium (Bartold et al., 2000). Fibroblasts, as the major cell population of the connective tissue, play a central role in formation, maintenance and repair of the connective tissue. Fibroblasts, in vivo, were once thought to be a homogeneous cell type regardless of tissue of origin, but it is now appreciated that considerable fibroblast heterogeneity exists and that numerous fibroblast subpopulations may exist within a given region (Bartold et al., 2000; Hakkinen et al., 1996; Zhou et a!., 2009). The concurrent residence of numerous periodontal fibroblast subpopulations, each with potentially different functions and expression patterns, has been implicated as a significant factor in the generation and maintenance of the periodontium, periodontal pocket development and formation of fibrotic gingival lesions (Hakkinen et al., 1996; Hakkinen and Csiszar, 2007; Lekic et al., 1997). Fibroblasts may 6 provide ECM factors or develop specific fibroblast subpopulations that affect the state of keratinization of the overlying epithelium, as there is ample evidence that epithelial keratinization is strongly influenced by the subjacent connective tissues (Karring et al., 1975; Ouhayoun et al., 1988). Accordingly, the non-keratinized connective tissues of “normal” junctional epithelium and sulcular epithelium are associated with sub-clinical levels of inflammation while the connective tissue subjacent to the oral gingival epithelium (keratinized) is not inflamed, an observation that is further supported by the subsequent keratinization of sulcular epithelium following experimental periods of intense oral hygiene (Nanci and Bosshardt, 2006; Page and Schroeder, 1976). Other histologic studies have observed that local fibroblast phenotype is dramatically affected during bacterially mediated gingival inflammation (Page and Schroeder, 1976). 1.1.3 Pocket Epithelium Pocket epithelium is histologically quite distinct from junctional epithelium or sulcular epithelium. Clinically, pocket epithelium is usually associated with increased sulcular depth though this is not a reliable distinguishing factor clinically as the primary differences are histologic and ultra-structural (Vitkov et al., 2005). Histologically, pocket epithelium shows marked proliferation of rete ridges, inflamed connective tissue papilla, bacterial accumulations and micro-ulceration, in stark contrast to junctional epithelium (Muller-Glauser and Schroeder, 1982; Vitkov et al., 2005). As an epithelial barrier, pocket epithelium is potentially less protective due to its reduced thickness, micro ulcerations as well as in vitro observations of reduced viability and proliferation of pocket epithelium cells (Bosshardt and Lang, 2005; Carro et al., 1997; Hunter et al., 7 2001). It must be clarified that the presence of the pocket epithelium does not preclude the presence ofjunctional epithelium. The junctional epithelium is still present, all be it at a significantly reduced dimension (Muller-Glauser and Schroeder, 1982; Overman and Salonen, 1994). As the junctional epithelium tends to migrate apically in periodontal pocket formation, there is a greater trend for the histologic hallmarks of pocket epithelium to be noted in the coronal half of the histologic sulcus (Muller-Glauser and Schroeder, 1982). Histologically, sulcular, junctional and pocket epithelium are non keratinized, however, the sulcular epithelium shows a tendency to keratinization, normally a sign of advanced cellular differentiation (Muller-Glauser and Schroeder, 1982). Junctional epithelium and pocket epithelium both are relatively undifferentiated, lacking ultra-structural signs of cellular differentiation such as keratohyalin granules or keratin filaments (Overman and Salonen, 1994). The undifferentiated state of these tissues has been speculated to be a cellular adaptive consequence of the local inflammatory process often initiated by the local bacterial plaque (Muller-Glauser and Schroeder, 1982). 8 1.2 Inflammation in the Periodontium 1.2.1 Bacterial Plaque Bacterial plaque can form on all surfaces of erupted teeth and generally develops in a predictable fashion when unimpeded by oral hygiene. Immediately following tooth prophylaxis, salivary mucins attach to the enamel surface creating an acquired pellicle which is promptly followed by attachment and supra-gingival colonization of gram positive cocci that provide the foot hold for further bacterial attachment (Theilade et al., 1966). Within a couple of days, gram-negative anaerobic bacterial begin to flourish, soon followed by the appearance of filaments and fusiform bacteria. These groups soon become the dominant species within the plaque biofilm. After one week the presence of spirochetes and vibrio species becomes apparent associated with the clinical onset of gingivitis (Loe et al., 1965). Concurrent with the development of this complex supra gingival biofilm, the plaque begins its growth apically into the gingival sulcus where it has a decidedly more gram-negative anaerobic profile. Periodontal disease has been associated with specific sub-gingival bacterial species, most notably Porphyromonas gingivalis, (Pg) Treponema denticola (Td) and Tannerella forsythia (Ti), most of which are anaerobic non-motile species. Successful periodontal treatment is associated with significant reductions of these bacteria (Haffajee et al., 2006). 1.2.2 Gingival Inflammation-Acute Chronic plaque accumulation along the dento-gingival junction presents a persistent bacterial challenge to the host. Innate host defenses include the continual shedding of epithelial cells and the continuous flow of gingival crevicular fluid into the sulcus and the 9 continual flow of neutrophils into the sulcus, all of which physically impede bacterial advances and are usually sufficient to maintain the homeostasis of the bacterial/host interaction (Madianos et a!., 2005). Interestingly, in bacteria-free animals neutrophils are histologically seen to be migrating through the junctional epithelium without any of the other signs of inflammation, suggesting that at some level, neutrophil migration in the junctional epithelium may be a physiologically normal process (Page, 1982). Occasionally when virulent bacteria, excessive plaque accumulations or a compromised host defense, upset the gingival tissue inflammatory balance the bacterial load will elicit a greater inflammatory response. Host cells recognize bacteria through highly conserved bacterial motifs called pathogen-associated molecular patterns (PAMPs), which are recognized by a subset of host cell receptors known as pattern recognition receptors (PRR). Numerous PAMPs exist, including lipopolysaccharides (LPS), peptidoglycans (PGN), lipotechoic acid (LTA) and fimbriae. PAMPs can be recognized by many cells but in the dento-gingival regions are usually first recognized by epithelial cells and the dendritic cells associated with the epithelium (Madianos et al., 2005). Subsequently these cells will release/express cytokines and chemokines. These chemical mediators of inflammation will function on the vasculature to increase local recruitment of more inflammatory cells and direct these cells to the site of inflammation. The acute inflammatory response effectively and rapidly develops a classic inflammatory response characterized by swelling, redness, heat and pain under the direction of these initial cytokines. 10 1.2.3 Gingival Inflammation-Chronic In a similar fashion, bacterial invasion or diffusion of PAMPs through the junctional epithelium will result in stimulation of connective tissue-associated cells such as resident macrophages, fibroblasts and mast cells, all of which will also release a series of cyokines, chemokines, prostaglandins, and matrix metalloproteinases (MMPs). MMP’ s released by fibroblasts will function to break down the extra-cellular matrix, primarily collagens, creating more space for the influx of serum and inflammatory cells. Resident macrophages and fibroblasts will secrete monocyte chemo-attractant protein-i (MCP- 1) that stimulate the recruitment of monocytes which ultimately differentiate to macrophages locally (Tonetti et al., 1994). Macrophages function primarily in phagocytosis, antigen presentation and immuno-modulation of the inflammatory response and signal the beginning of the chronic inflammatory response (Fujiwara and Kobayashi, 2005). Chronic inflammation, mediated by a whole host of pro-inflammatory and pro- resolving cytokines is associated with lymphocyte recruitment, blood vessel proliferation, and tissue fibrosis. If the offending bacteria are removed the inflammatory response will resolve and macrophages will be involved with tissue repair and removal of apoptotic cells (Kantarci and Van Dyke, 2005; Seymour and Gemmell, 2001). If the bacterial challenge remains, chronic inflammation will ensue, extending the range and nature of the inflammatory response. The persistence and subsequent maturation of the sub-gingival plaque biofilm will bring into play the adaptive immune response under the influence of the macrophages and the accumulating lymphocytes and their combined milieu of locally generated cytokines. 11 Antigen presentation, primarily by dendritic cells and macrophages, to naïve T-cells is critical to the adaptive response as the immune system attempts to refine it’s attack specifically against the offending bacteria. The result is an increase in local lymphocyte accumulations under the direction of locally secreted cytokines (Jenkins et al., 2001). Up till this point the lesion generally remains localized to the gingival tissues, however, if a local change in susceptibility occurs, possibly mediated by local infection, changes in granulocyte function, changes in T-cell profile or systemic factors such as disease or stress, then irreversible involvement of the supporting tissues may ensue (Page et a!., 1997). Histologically, impending attachment loss is linked to a B lymphocyte mediated lesion, dominated by plasma cells (Gemmell et al., 2002; Page and Schroeder, 1976). The connective tissue breakdown that had been previously limited to the gingival tissues, extends to the alveolar bone and periodontal ligament resulting in irreversible loss of attachment (Gemmell et al., 2002; Page and Schroeder, 1976). Ultimately, the local cytokine profile, modulated by both known and unknown risk factors will determine if the attachment loss and crestal alveolar bone loss, that characterize periodontal disease, develop (Page et al., 1997). 1.2.4 Mediators of Inflammation-Cytokines Cytokines are low molecular weight polypeptides secreted by many cell types, but primarily ascribed to macrophages and lymphocytes, that modulate the activity and production of various effecter cells (Seymour and Gemmell, 2001). Cytokines bring about their outcome by effecting cell surface receptor activation and expression, work synergistically with or antagonistically against other cytokines, and have the ability to 12 produce apparently opposite actions when complimented together in different cytokine profiles. As discussed above, the B-cell mediated lesion is generally associated with tissue breakdown while the T-cell mediated lesion reflects stability, and the balance between a T-cell dominated lesion and a B-cell dominated lesion is regulated by the T helper cells (Th) (Berglundh and Donati, 2005). Immature T-helper cells are presented antigen and depending on the manner the antigen is presented, they mature from naïve T cells into T-helper cells. T-helper 1 cells (Thi) secrete primarily interleukin (IL) -2, tissue necrosis factor-a (TNF-a) and interferon-y (INF-y) promoting a cellular immune response by T-lymphocytes. T-helper 2 cells (Th2) secrete primarily IL-4, IL-5, IL-6, IL- 10 and IL-13, and promote a humoral immune response mediated via antibody production from appropriately activated B-cells (De Martinis et al., 2006; Jenkins et al., 2001; Piccinni, 2006; Robertson and Hansson, 2006). Although it is unclear how naïve T cell differentiation is modulated towards Thi or Th2, though local cytokine profiles seem to be important in this determination, the ratio of the Thl :Th2 cytokines is probably more critical than the absolute presence of one profile (Bonecchi et al., 1998; Szabo et al., 2003). Consequently, what seems well identified is periodontal disease is associated with elevated levels of pro-inflammatory cytokines intereukin- 1 l (IL-i Ia), INF-y, prostaglandin E2 (PGE2), tissue necrosis factor alpha (TNFa) and MMP. Conversely, healthy gingiva is associated with high levels of IL- 10, tissue inhibitor of metalloproteinases (TIMPs), and transforming growth factor betal (TGFf31) (Dutzan et al., 2009; Page et al., 1997). 13 Transforming Growth Factor Transforming growth factor beta is considered a prototypical member of a large group of proteins in the TGF beta family that includes bone morphogenic proteins, activins and growth differentiation factors (Brown et al., 2007; Li et al., 2006). TGFI31 is the most common of the three mammalian isoforms, the others being TGFI32 and TGFI33. TGF3l primarily functions to control ECM development, angiogenesis, and the initiation and resolution of inflammation by regulation of lymphocyte proliferation, differentiation and survival (Li et al., 2006; Prime et al., 2004). The three forms activate similar receptors and produce similar intracellular signaling pathways, though different clinical effects seem to be associated with each isoform (Eslami et al., 2009; Schrementi et al., 2008; Sheppard, 2006). The importance of TGFf31 as a potent dampener of the inflammatory process is highlighted by the observation that TGFI31 deficient mice show early mortality associated with diffuse mono-cellular infiltrations (Kulkarni et al., 1993). TGFI31 is stored in the ECM and is secreted in a latent form, providing a quick ample supply when needed but protected from inadvertent activation. During intracellular assembly prior to secretion, TGFI31 is non-covalently bonded to its latency-associated peptide (LAP), the unit conjointly being referred to as the small latent complex (SLC). Further intracellular modifications often create disulfide bonding between the LAP and latent TGF3 binding protein (LTBP), creating a large latent complex (LLC) (Lamers et al., 2007). Non-covalent binding of TGFf31 with the LAP ensures TGFI3’s are secreted in an inactive form and if associated with the LTBP, the TGFI3 1 is secured to the ECM, by fibrilin-1 and fibronectin (Dallas et al., 2005). LTBP are a family of fibrin-like molecules 14 that regulate bioavailability and activation of TGF(3 1. Three of the family of four known LTBP-members, LTBP-i, LTBP-3 and LTB-4 are covalently bound to the SLC while LTBP-2 does not show evidence of binding to the SLC (Zhou et al., 2009). ECM storage of TGFI3 1 is common amongst healthy mammals, with little evidence of inherent activity unless directly activated (Sheppard, 2006). Activation of latent TGFI31 occurs by three known mechanisms. Proteolytic activation has been identified in vitro by the serine protease, plasmin, as well by the matrix metalloproteases, MMP-2 and MMP-9 (Annes et al., 2003). Non-proteolytic activation of TGFII1 as been proposed by the extra-cellular matrix protein thrombospondin-1 (TSP 1) which oxidatively alters the affinity to TGFI31 via binding to a N-terminus conserved sequence of LAP that is critical to confer latency to TGF(1 within the SLC (Aluwihare and Munger, 2008; Young and Murphy-Ullrich, 2004). The importance of TSP-i towards TGFt31 activation has been observed with phenotypic similarities of TSP-l null mice pups are similar but to a lesser extent, to TGFI31 null mice that can be rescued by exogenously administered TSP-l-derived peptide that activated TGF-31 (Crawford et al., 1998). Finally, integrin av136 binding to the arginine-glycine-aspartic acid (RGD) sequence on LAP of the ECM bound LLC of TGFI31, purportedly, induce temporospatial alterations in the association of the av and f36 subunits (Luo et al., 2007). These changes intra-cellularly initiate a cellular cascade that ultimately generates an actin-mediated traction to the LLC via av136 integrin that disrupts the weak electrostatic forces between the TGFI3 and the LAP, releasing TGF31 (Harburger and Calderwood, 2009; Keski-Oja et al., 2004). Recently a fourth mechanism of TGFI31 activation has been reported where 15 by reactive oxygen species, specifically hydroxide, could induce a conformational change in LAP-f31, resulting in the release of TGFI31 (Jobling et aL, 2006). This observation allows TGFf31 to be rapidly released subsequent to conditions releasing or producing local ROS, like neutrophil de-granulation or ionizing radiation, respectively. Liberated TGFI31 mediates its cellular effects by binding to a group of trans-membrane proteins called TGFI31 receptors (TR). TGFfH binding to TI3R-II leads to assembly with TI3R-I forming a hetrotetrameric complex (Itoh and ten Dijke, 2007; Prime et a!., 2004). The activated complex formation becomes internalized and allows intracellular activation of a group of intracellular proteins called Smad proteins by either clathrin dependant pits or lipid raft caveolae formation, each with dramatically different results. Clathrin dependant internalization involves R-Smad binding, ultimately leads to nuclear translocation where these proteins bind to specific promoter regions on target genes regulating expression of target genes (Itoh and ten Dijke, 2007). However, involvement of 1-Smads is associated with lipid raft caveolae which precludes nuclear translocation and no subsequent expression of target genes (Itoh and ten Dijke, 2007). Intracellular receptor trafficking in this manner is one mechanism for regulation of TFGI3 1 activity. While in vitro TGFI31 activation has been achieved by numerous different mechanisms, biologic significance of these has not be identified (Sheppard, 2006). In-vivo TGFI31 activation has been confirmed by binding with plasmin, reactive oxygen species, thrombospondin- 1, integrin av36 and integrin av138 (Schrementi et al., 2008; Yang et a!., 2007) while in vitro, binding of integrin ccvI35 by myofibroblasts has induced activation of latent TGFI31 (Wipffet al., 2007). 16 1.3 Interin avB6 Integrins are a family of trans-membrane proteins that are involved with the mediations of cell adhesion, growth, homeostasis and differentiation (Luo et al., 2007; Watt, 2002). They are heterodimeric trans-membrane proteins integral in cell-cell interactions and cell ECM interactions and play important roles in tissue morphogenesis, remodeling, repair, guided cell migration and formation of epithelial cell barriers (Humpbries et al., 2006; Hynes, 2002; Jones and Walker, 1999; Uitto and Larjava, 1991). Integrins present as a pairing between a a-subunit (18 known subunits) and a 13-subunit (8 known subunits) that are non-covalently associated with each other (Berrier and Yamada, 2007; Breuss et al., 1995; Hynes, 2002; Sheppard, 2000). Twenty-four integrin heterodimers have been identified to date. The integrins function as trans-membrane proteins with large extra- cellular domains and small intracellular domains, excepting the (34 integrin subunit which has a larger intracellular region of about 60 amino acid residues at the carboxy (C-) terminus (Humphries et al., 2006; Thomas et al., 2006). Integrins were initially thought to function primarily as binding sites for extra-cellular molecules, however more recent research has confirmed that these molecules also function significantly as cell signaling molecules (Harburger and Calderwood, 2009; Jones and Walker, 1999). These signaling actions can be modulated from within the cell to affect the affinity state of extra-cellular binding (inside-out signaling), or from the extracellular matrix (ECM) into the cell, transmitting critical information on the cell’s immediate environment (outside-in signaling) (Berrier and Yamada, 2007; Jones and Walker, 1999). Integrin molecules have no intrinsic catalytic activity and as such must bind to accessory molecules that 17 cooperatively can impart a local effect (Legate et al., 2009). Integrin signal transduction represents a large body of current research activity. Recent mapping of the network of intracellular molecular interactions associated with the integrin adhesion complex has illustrated the depth and breadth of the complexity of integrin signaling (Harburger and Calderwood, 2009). This mapping has identified 156 different components (molecules, lipids and ions) being linked by 690 potential configurations. The many redundant interactions and complementary pathways is thought to play a critical role in making the integrin adhesion complex resilient to failures at individual links (Zaidel-Bar et al., 2007). Integrins are often expressed on cell surfaces in inactive forms, in essentially low-affinity binding state. Cellular activation of integrins typifies inside-out signaling (Luo et al., 2007). Studies of molecular stereochemistry suggests that this low-affinity binding state is characterized by a bent conformation of the extracellular integrin domains and a close approximation of both dimers through the plasma membrane and including their intracellular tails (Harburger and Calderwood, 2009; Luo et al., 2007). Inside-out signaling is achieved when an intracellular signaling cascade associated with the short cytoplasmic tails induces a conformational change in the orientation of the a and 13 subunits (Calderwood, 2004). The binding of talin to the cytoplasmic region of the f3- subunit has been well established as a central molecule in the integrin intracellular activation cascade (Harburger and Calderwood, 2009). Recent findings also point to the kindlin proteins as critical modulators of inside-out integrin signaling (Harburger et al., 2009; Larjava et al., 2008). Though these two molecules seem to have central roles in 18 inside-out integrin signaling, they are by no means the only ones involved. The conformation change in the integrin heterodimer has been described as a swinging-out movement of the extracellular domain, likened to a switch-blade motion associated with a separation of the two integrin subunits throughout their full length such that the extra- cellular binding site is exposed, facilitating binding to the intended extracellular ligand (Luo et a!., 2007; Zaidel-Bar et al., 2007). Outside-in signaling is an equally complex process initiated when integrins bind to their ECM bound ligands. Intracellular effects involve immediate down stream activation of lipid kinase activity, increasing concentrations of phosphoinositide second messengers and phosphorylation of integrin bound molecules within the intracellular adhesion complex, resulting in a robust activation of many classes of signaling molecules including the RhoA guanosine triphosphoesterases (GTPases), extracellular-signal regulated kinase (ERK) pathway and Jun N-terminal kinase (JNK) pathway (Zaidel-Bar et a!., 2007). Subsequent biologic responses following integrin binding include gene expression, cell cycle regulation, focal adhesion turnover and actin dynamics (Legate et a!., 2009). Though the consequences of extracellular integrin binding represents a broad range of physiological actions in both differentiated and undifferentiated cells, a common factor seems to involve cytoskeletal reorganization and the controlled application of force via the actin intermediate filaments (Harburger and Calderwood, 2009; Legate et al., 2009). Interestingly, it has been recently reported, in vitro, that matrix stiffness has the potential to dramatically alter mesenchymal stem cell (MSC) differentiation in otherwise identical conditions, suggesting the functional necessity of a traction force to sense the 19 nature of the ECM which acts as a co-determinant of the subsequent signaling cascade (Engler et al., 2006). Considering the limited number of integrin heterodimers and the extensive catalogue of cellular activities they are involved with, the modulation of the intracellular signaling based on physicial properties of the ECM may, in part, account for the complexity of the integrin adhesion complex, effectively increasing the range of actions a particular integrin heterodimer may be responsible for. Integrin avf36 is an epithelial associated integrin. Integrin av36 is not expressed in most adult differentiated epithelial tissues, but is induced soon subsequent to wounded or injured epithelia as well as in various cancers, including colon, cervical, lung, breast and oral carcinoma (Abmed et al., 2002; Berrier and Yamada, 2007; Breuss et al., 1995; Thomas et al., 2006; Zambruno et al., 1995). The extra-cellular domain of integrin av6 binds to the arginine-glycine-aspartic acid (RD0) site of the extracellular matrix proteins tenascin (Prieto et a!., 1993), osteopontin (Huniphries et a!., 2006), fibronectin, vitronectin (Huang et al., 1998a) and LAP1 and LAP3 of latent TGFI31 and TGF3 respectively (Annes et al., 2002; Annes et al., 2004; Humphries et al., 2006; Munger et al., 1999; Prieto et al., 1993). Fibronectin was identified early on as a ligand for integrin av136 (Breuss et al., 1995). Protein kinase C (PKC) was one of the first enzymes associated with av(36 integrin functionality and identified as central to its functionality as inhibition of PKC reduced fibronectin and vitronectin mediated migration. This finding is substantiated by PKC holding a central hub of interaction in the integrin adhesion complex (Huang et al., 1998a; Zaidel-Bar et al., 2007). The association of fibronectin with the provisional matrix expressed during early wound healing stimulated interest in 20 how av136 integrin may be involved in this process. Integrin av[36 expression was evaluated and determined to be undetectable in unwounded epithelium and during early wound healing but expressed following confrontation of the migrating keratinocytes with expression maintained during the phase of epithelial re-organization and re-establishment of the basement membrane after which its expression is reduced to pretreatment levels (Breuss et al., 1995; Hakkinen et al., 2000). The association of av36 integrin with carcinoma as discussed above is well established and seems to underscore this integrin’ s association with tissue re-organization (Haapasalmi et al., 1996; Thomas et al., 2006). Significant interest lies in the association of ccvI36 integrin expression in carcinoma keratinocytes with the metastatic propensity of the tissue. In studies of oral squamous cell carcinoma, integrin av136 was identified in 73-100% of the biopsy samples while an examination of biopsy tissue associated with progressing oral leukoplakia lesions, all samples were positive for staining to av36 integrin (Hamidi et al., 2000; Jones et a!., 1997). Integrin av136 has been associated with inhibition of anoikis, and suggested as a potential mechanism by which epithelial tumor cells can grow in the absence of attachment to basement membrane and leave their immediate environment potentially allowing the tissue to acquire a metastatic potential (Janes and Watt, 2004). Recently, using cells cultured from human colon cancer samples, it has been reported that levels of avf36 integrin correlate positively with secretion of pro-MMP-9 and pro-MMP-2, which when activated could play a significant role in degradation of the ECM, a necessary requirement of an invasive carcinoma (Wang et al., 2008). 21 1.3.1 Genetic Knockout Studies Genetic knockout studies have contributed extensively to current knowledge regarding integrin av136. The av integrin subunit can combine with 5 different 13 integrin subunits (131, I3, 135, 136 and 118) while the (36 integrin subunit is only known to naturally pair with the av subunit (Humphries et al., 2006; Hynes, 2002). Owing to the singular known association of the (36 integrin with the cv integrin, the homozygous 136 genetic deficient mouse strain (136-I-) is effectively an av136 knockout. (Huang et al., 1996; Sheppard, 2006). The 136-I- mouse strain is viable and seems to live and reproduce normally though they express significantly greater skin and lung inflammation as a consequence to local trauma in comparison to wild-type littermates. Interestingly, even though av136 integrin is normally expressed during wound healing, no deficits were noted to skin wound- healing experiments in these knockout mice (Huang et al., 1996; Sheppard, 2006). Further investigations noted that the lung inflammation could be effectively prevented by a transgenic rescue strain constituently expressing the 136 integrin, thus definitively attributing the effect of increased inflammation to the absence of the 136 integrin sub-unit (Huang et al., 1998b). The finding that lung tissues of the 136-I- mouse strain were resistant to bleomycin induced post inflammatory fibrosis ultimately lead to the observation that avf36 integrin was directly associated with TGF131 activation (Munger et al., 1999). In vitro studies presented in the same publication suggested that TGF13-activation required BOTH the extra-cellular binding of integrin av136 to TGFf31 LAP1 as well as intracellular binding of 22 6 integrin sub-unit to a functional cytoskeleton. (Munger et al., 1999; Sheppard, 2006). As previously mentioned, all known effects of av6 integrin are mediated via binding with the RGD amino acid sequence in its target protein. Similarly, activation of TGFI31 by avj36 integrin involves binding but the target protein is LAP1 which as has been confirmed by in vivo studies where a mutation of the binding sequence to argine-glycine glutamic acid (RGE) eliminated TGF31 activation. (Annes et al., 2004) Integrin av6 is strongly associated with regulation of TGFI3 activity, as experimentally blocking its ability to bind to the LAP has been shown to induce a phenotype showing strong similarities to TGF3 knock-out mice (Yang et al., 2007). The binding of av136 integrin to LAP 1 coupled with cellular traction, facilitates the release and subsequent activation of TGF1 (Annes et al., 2004) (Sheppard, 2001) Further supporting it’s role in reducing inflammation and promoting fibrosis, the presence of integrin av136 has been associated with tissue fibrosis. In human biopsies of fibrotic kidney tissues associated with various disease states, the presence of integrin av136, determined by immunohistochemical analysis, was strongly associated with kidney fibrosis (Hahm et al., 2007). Experimental induction of kidney fibrosis using a model of unilateral urethral obstruction, revealed that 136-I- mice were subject to less kidney fibrosis (Ma et al., 2003). Similarly, examination of human lung tissues biopsies has shown that level of fibrosis strongly correlates to the presence of integrin av136, as determined by immunohistochemical testing (Horan et al., 2008), while as previously discussed, 136-I- mice are protected from bleomycin induced lung fibrosis (Munger et al., 1999). 23 1.3.2 Integrin av6 and the Junctional Epithelium Owing to the reputed anti-inflammatory properties ascribed to TGF31 discussed above, recent investigations have sought to identify if the dentogingival junctional epithelium may be a candidate for expression of ccvi36 integrin. A recent investigation identified that deficiency of integrin av136 was linked to the initiation of periodontal disease (Ghannad et a!., 2008). In their assessments of clinically healthy human gingiva as well as mouse and rat gingiva from health animals, they definitively showed expression of integrin av(36 to all layers of the JE, co-localizing with the distribution of TG931, while an examination of human gingival tissues associated with advanced periodontal disease showed a dramatic down-regulation of integrin av36. Histologic gingival examination of 136-I- mice revealed significantly greater apical epithelial migration compared to age- matched wild-type groups (Ghannad et al., 2008). In an attempt to associate TGFI31 activation with integrin av136, the same group reported a likewise greater epithelial migration in a rat model of LPS-induced periodontal disease by functionally blocking av136 integrin activation of TGFI31. To evaluate the potential anti-inflammatory properties of TGF(31, an assessment of JE inflammation revealed that 136 integrin deficient mice strains showed significantly greater inflammatory scores than wild type control groups, supporting the contention that JE activation of TGF131 by av136 reduces local inflammatory stress. Finally, in at attempt to quantify the amount of alveolar bone loss associated with integrin 136 deficiency, the same group reported that 3-month, 6- month, and 12-month-old groups of 136 integrin deficient mice showed significantly greater alveolar bone loss at the first and second molars compared to age-matched wild- type groups. Constitutive expression of integrin av136 and its activation of TGFI31 in the 24 healthy junctional epithelium, may help offset the chronic inflammatory stress of bacterial plaque there by stabilizing the junctional epithelium and explain why the 136-/ strain of mice may exhibit greater levels of gingival inflammation and the associated accelerated alveolar bone loss. 1.4 Diabetes Mellitus and Periodontal Disease Periodontal disease has many known risk factors, and notably amongst those risk factors is a group of diseases clinically associated with chronic hyperglycemia, collectively referred to as diabetes mellitus (Mealey and Ocampo, 2007; Taylor and Borgnakke, 2008). Diabetes mellitus often clinically presents with a varied combination of polydypsia, polyphagia, polyuria, weight loss, fatigue and/or weakness, which have been classically used as cardinal signs of poor glucose management. The association between diabetes and periodontal disease has been well established so much so that periodontal disease has been proposed as the sixth complication of a diabetes mellitus (Loe, 1993). A fasting plasma glucose level 126 mg/dl (7 mmol/L) that is confirmed with a second test on a different day confers a diagnosis of diabetes mellitus (Alberti and Zimmet, 1998). Treatment of this group of diseases aims at a stable reduction of the levels of blood glucose levels below 100 mg/dl (5.5 mmolJL), and may involve oral hypoglycemic agents, insulin injection, modification of lifestyle or a combination of theses modalities (Mealey and Ocampo, 2007). Measurement of glycated hemoglobin using the hemoglobin (Hb) Al c test, is usually used for monitoring long-term glycemic control, with goals for diabetic patients to keep these levels below 7%, while non-diabetic patients are usually below 6% (Alberti and Zimmet, 1998). As a risk factor for periodontal 25 disease, numerous studies have found that diabetic adults and adolescents experience greater incidence, prevalence of periodontal attachment loss and more severe disease compared to control groups, the increased risk ranging from 1.8x to 3.7x depending on the measures used to define disease (Lalla et al., 2007; Saito et al., 2004). A high degree of glycemic control is associated with protection from continued periodontal attachment loss and is associated with no significant differences in outcomes in periodontal therapy between diabetic and control groups (Seppala and Ainamo, 1994; Westfelt et al., 1996). While evidence generally supports the effect of uncontrolled diabetes on periodontitis, recent observations suggest that periodontal disease can also affect glycemic control. Reductions in HbAlc levels with non-surgical therapy have been consistently reported in the 0.5-1.0% ranges over the 3-6 month periods following therapy (Faria-Almeida et al., 2006; Grossi et al., 1997; Kiran et al., 2005). Although an exact mechanism linking periodontal disease and diabetes mellitus has not yet been elucidated the actual mechanism may represent a multi-factorial inter relationship (Nelson, 2008). Insulin resistance is increased during acute bacterial infections by TNFa which suppresses insulin-mediated skeletal muscle uptake of glucose, further increasing blood glucose levels (Grossi and Genco, 1998). Bacterially induced inflammation is significant in untreated periodontal disease secondary to the large burden of periodontal pathogens, a situation that is further compounded by the relatively thin and ulcerated pocket epithelium (Mealey and Oates, 2006). Chronic hyperglycemia leads to accumulation of non-enzymatic glycation of proteins and lipids in the plasma and tissues, referred to as advanced glycation end-products (AGE) (Lalla et 26 al., 1998b). AGEs increase ECM cross-linking, impeding tissue turnover and maintenance and significantly affecting cytokine release from inflammatory cells (Acosta et al., 2008; Lalla et al., 1 998b). Increased levels of AGEs have been identified in tissue biopsies from diabietic patients undergoing periodontal surgery compared to non-diabetic patients undergoing similar procedures (Lalla et al., 1998b). Additionally, AGE can up- regulate the expression of its own receptor, further amplifying the local effects of AGEs (Katz et al., 2007). Receptors for AGE (RAGE) have been identified associated with the small vessels of the lamina propria and the stratum basalis and spinosum in the gingival tissues of normal and diabetic patients with periodontal disease (Katz et al., 2005). Furthermore, RAGE has been identified on the surface of monocytes and macrophages where it can induce apoptaxis (reduced motility) and expression of IL-i f3 and TNFa messenger ribonucleic acid (mRNA). (Goldin et al., 2006; Mealey and Ocampo, 2007) On endothelial cells, RAGE-AGE binding will increase permeability increasing tissue edema (Goldin et al., 2006). In vivo studies using a peritonitis model have shown that diabetic mice have significantly elevated levels of leucocyte recruitment compared to control animals (Chavakis et al., 2003). On endothelial cells RAGE mediates an increase in classic endothelial cell adhesion molecules (ICAM- 1, VCAM) as well as acts as a direct receptor to the MAC-i receptor on the neutrophil surface (Chavakis et al., 2004). Fibroblast RAGE-AGE binding will result in decreased collagen synthesis, tipping the balance towards tissue breakdown (Lalla et al., 2000; Niu et al., 2008). Highlighting the pro-inflammatory capacity of RAGE-AGE interaction, animal studies employing a blockade of RAGE with soluble RAGE (sRAGE) reported a reduction in alveolar bone loss in a dose dependent manner, concomitant with reductions in pro-inflammatory 27 cytokines TNF-cx, IL-6 and MIvIPs (Lalla et al., 2000). The net periodontal effect of hyperglycemia, is an aggravation of gingival as well as systemic inflammation. The presence of periodontal attachment loss is associated with increased risks of various systemic diseases in addition to diabetes, though causation has not been conclusively established in these diseases either (Jin et al., 2003). Associations between periodontal disease and cardiovascular diseases and adverse pregnancy outcomes, have been recently proposed as related to the effects of periodontally generated cytokines (Loos, 2005; Offenbacher et al., 1996). Periodontal disease, as mediated by dental accumulations of bacterial plaque, is characterized by local inflammation that unpredictably results in periodontal attachment loss, however, the study of the potential systemic inter relationships is problematical when periodontal disease initiation and progression is unpredictable (Kinane and Attstrom, 2005; Page and Schroeder, 1976). Various animal models of periodontal disease have been established to facilitate the study of the patho physiology of periodontal disease, including ligature induced periodontal disease and LPS induced periodontal disease (Graves et aL, 2008; Page, 1982). Associations between periodontal disease and uncontrolled hyperglycemia have encouraged the development of a model of periodontal disease using diabetes as a means of increasing inflammatory burden. The results in a murine model confirmed the elevated levels of alveolar bone loss and periodontal breakdown capacity (Lalla et al., 1998a). As discussed above, uncontrolled diabetes mellitus represents a hyper-inflammatory state with increases in pro-inflammatory cytokines and chemokines (Mealey and Ocampo, 2007). Intra peritoneal injections of streptozotocine have been commonly used to induce a 28 6t (9L6T‘!U!SS0}IPW0)111J66T‘1PUW1puiaixna)sjpo-çjoJjEonuPdalpisuu ICTiunwuJ!0nnpaonpupunpunspo-jonanundaipJOUOflOflhlSOP‘Pfl1rn’-’”°-’! unjouonnuiqwoonijnosqjsalaqnipjac1Xunpiu.nwiCq031wU!ainisoiuiaoXjsadXq 1.5 Hypothesis, Aim and Objectives of the Study The gingival epithelium is an integral part of the innate immune system, functioning against the local invasion of periodontal pathogens and towards maintaining tissue homeostasis (Teng, 2006). Recent observations have suggested that integrin av(6 activation of TGF(31 by the junctional epithelium mediates an anti-inflammatory effect thereby conferring a measure of protection to periodontal disease (Ghannad et al., 2008). Uncontrolled diabetes imparts a significant systemic inflammatory response that is also measurable periodontally (Lalla et a!., 2001). The hypothesis for this study is that experimental diabetes, as an independent pro- inflammatory factor, will further exaggerate periodontal disease in 136 knock-out mice, while 136 over-expression will protect mice from periodontal disease and reintroduction of this integrin under a different genetic promoter will rescue mice from periodontal disease. Using a strain of previously characterized 136 knock-out mice (Ghannad et al., 2008), we aim to show that, relative to non-diabetic mice, introduction of experimental diabetes will increase murine periodontal disease, as measured by periodontal inflammation, alveolar bone loss and gingival migration. To reinforce the periodontally protective role of av136 integrin, we hope to show that a previously reported (Hakkinen et a!., 2004) transgenic murine strain, with human 136 integrin (h136) over-expression under the cytokeratin 14 gene promoter (K14136), will be protected from periodontal disease. As further evidence of integrin av136 protection from periodontal disease, a novel genetic rescue strain, 30 developed by crossing the 136 over-expressing and 136 deficient strains ((36 Rescue), will re-establish the levels of periodontal disease activity seen in W/T mice. In a murine model following to the induction and maintenance of uncontrolled experimental diabetes in four groups of mice (WIT, (36-I-, 136 Rescue and K14136) and parallel with matched non-diabetic mouse groups, all animals will be histologically and histomorphometrically analyzed with a three-fold objective; 1. To measure amount of gingival epithelial migration along the molar tooth surfaces as a measure of periodontal instability. 2. To quantify the area alveolar bone loss at the molar teeth as a measure of irreversible periodontal disease. 3. To grade the level of gingival leukocyte infiltrate at interproximal molar sites as a measure of gingival inflammation. 31 CHAPTER 2: MATERIALS AND METHODS 2.1 Induction of Experimental Diabetes Mellitus One experimental group from each of the following mouse strains; wild-type (WT), 136 deficient (136-!-), 136 overproducing (K14136) and a 136 rescue (136 Rescue) were selected for inclusion in the study. Each group consisted of five male and five female nine-month old mice, which were randomly selected from resident breeding colonies. Animals were individually weighed and subsequently rendered diabetic with 6 consecutive daily intra peritoneal streptozotocine (Sigma-Aldrich, St. Louis, MO, USA) injections (55mg/kg) diluted in sterile citrate buffer (0.5 M; 4.5 pH) that were freshly prepared immediately prior to use (Brosius, 2003). All animal manipulations and injections were performed under isoflurane (AErrane, Baxter Corp, Ontario, Canada) general anaesthesia. All animal procedures and housing were performed with prior approval of the University of British Columbia Animal Care Committee. Age and gender matched control groups with similar numbers of animals from each strain, were similarly injected with equivalent aliquots of the citrate buffer vehicle (0.5 M; pH 4.5). Three weeks following the last injection, all groups were fasted (4 hours), animals anaesthetized and tail blood samples were collected for blood glucose testing (Accu-Chek Compact - Roche Diagnostics, Quebec, Canada). The blood glucose monitor was tested prior to all animal manipulations against standardized solutions, as directed by the manufacturer. A blood glucose level of 306 mg/dL (17 mmol/L) was considered a confirmation of diabetes in the experimental group (Lalla et al., 2000). Experimental animals that failed to become diabetic were exposed to two more consecutive daily doses of streptozotocine at 32 p3wumpspqjowqojouopTo3oloadutoposodx1OM SjW!Lpjmpuptm‘uisipi&pojdai‘qssoduoqpupms oqwoijpAOWaIaipoudui.mp£pmwaidpipwqspnun uoiuunpsaLuTjouTssssuduu?puippoqpirn uo!dwrlsuo1OMpsJouspnssooqrnpJoSuiS1AOiojpusqoAjuoiiipp r”ppsIw!u1jj‘poudjmuwuodx91 .wqflpviowpui(ioocui.md)otposnowpmpus ossoqiikqizjU!Uqiu/ApJno1J-JiwposnoqaLsjiunuuv (‘TjjowtuLI>)‘jp,iw90E>SPAlsoonjpoojqufls1JAiqopuuguoosdnoi IIUO.IJSJUI!U11OJUOO‘SUOi1fUi&IIMOJJOJS)1N‘ATS1!W!SApmswoIj OAOWJ1Mmqipuioqopjnjimpj’vSpAJsoonjpooq.ioJposoi pu1pisJupaiSjUJtU1O1JS)OMOMJAJJUJIpsnsopuis 2.2 Tissue Preparation and Assessment Four months following confirmation of a diabetic state, fasted animals were anesthetized under general anesthesia, weighed, had tail blood samples evaluated for blood glucose levels, and then sacrificed in a C02 chamber. The animal was immediately decapitated and maxilla and mandible disarticulated with the aid of a light microscope (Leica MZ6 Switzerland, objective 4x, eyepiece lOx) for standardized preparation. Control animals were similarly treated following the 16 week experimental time period. Figure 1 presents a summary of the experimental protocol utilized in this study. W/T (36-I- (36 Rescue Randomized Groups (each sub-group 5 c?1 and 5 ) EJFZI, Control Experimental Group Streptozotocirie x6Days 3wks Tissue Examination QIMx Decalcify & Paraffin _____________ Md-De-flesh _____________ 2.2.1 Mandible The harvested mandibles were immediately immersed in 2% aqueous potassium hydroxide (Fisher Scientific, Fair Lawn New Jersey, USA) that was regularly changed as Vehicle Only’ x 6 Days Blood Glucose Test <306 mgldL 16 wks Sacrifice 16 wks \l1 4- Figure I. Experimental Protocol Sacrifice 34 necessary over a period of 4-6 weeks until samples were completely dc-fleshed. To facilitate identification of the cemento-enamel junction, de-fleshed samples were stained as previously reported (Ghannad et al., 2008). Briefly, all samples were stained using Van Gieson’ s solution (http://stainsfile.info/StainsFile/stainIconektv/vangies.htm) for 30 seconds immediately followed by immersion in a 1:10 aqueous dilution of Ponceau-S solution (Sigma-Aldrich, St. Louis, MO, USA) for 5 minutes. Stained mandibles were rinsed quickly in distilled water and dried completely prior to assessment. Quantification of Root Exposure and Crown Width Area of root exposure was quantified as described previously (Ghannad et al., 2008). Accordingly, stained and dried hemi-mandibles from all term animals were fixed on a movable platform with the lingual surfaces visible. The samples were oriented with the aid of a light microscope (Leica MZ6 Switzerland: objective 4x, eyepiece lOx) until the buccal and lingual cusps were superimposed and perpendicular to the line of site. Digital images were taken (Nikon, Coolpix 995, Tokyo, Japan) of the molars while under magnification (Leica MZ6 Switzerland: objective 4x) with a measurement scale for reference (PCP 126, Hu-Friedy, Chicago, IL, USA) and saved as Joint Photographic Experts Group (JPEG) files. Images were subsequently opened in LmageJ software (http://rsb.info.nih.gov/ij) and standardized for magnification using the known dimensions of the standardization scale in the digital images. The area visualized on the digital image between the CEJ and alveolar crest on the lingual surfaces of each of the first molar and second molar were circumscribed with the aid of a computer mouse and 35 then quantified using ImageJ software. Figure 2 illustrates an example of measuring the exposed root surface on the first molar. Figure 2. Quantification of the root surface area between the CEJ and alveolar crest of the first molar using lmageJ software (40x magnification). jaws have been stained to facilitate identification of the CEJ and alveolar crest Images taken with the camera lens oriented perpendicular to the root surface. All quantification was done at a constant magnification, by one examiner and reliability measurements were verified to be highly significant (Pearson correlation P-value=O.942). Using the same standardized digital image, the tooth width (distance between the mesial and distal contact) of each of the first and second molars were measured. All measurements were made by a single examiner (SA). At a different sitting, repeated measures on a random selection of the images were completed for both exposed root area and tooth width as an assessment of intra-examiner consistency. Figure 3 illustrates an example of measuring the width of the first molar. 36 Figure 3. First molar crown width measurement using lmageJ software (40x magnification). Jaws have been stained to facilitate identification of the CEJ and alveolar crest. Images taken with the camera lens oriented perpendicular to the root surface. All quantification was done at a constant magnification, by one examiner and reliability measurements were verified to be highly significant (Pearson correlation P-valueO.979). 2.2.2 Maxilla The hemi-maxilla (right side) was immediately fixed in 4% formaldehyde (Fisher Scientific, Fairlawn, NJ, USA) in PBS (Invitrogen-GIBCO, Carlsbad, CA, USA) for 2 days, then decalcified in 2% formaldehyde and 0.4M ethylenediamine tetracetic acid (EDTA) (Sigma- Aldrich, St. Louis, MO, USA) in Phosphate Buffered Saline (PBS) (Gibco, Grand Island, NY, USA), pH =7.2. Solutions were changed weekly until decalcification was confirmed radiographically (Insight-F speed dental radiographic film, Eastman Kodak, Rochester NY, USA at 15 mA, 50 mV, 0.8 see, General Electric 100 Radiographic Unit, Milwaukee WI, USA) by uniform radio-opacity of the exposed radiographic images. Decalcified hemi-maxilla were dehydrated in ascending alcohol solutions and finally in Xylene (Histoprep, Nepeane, ON, Canada) prior to embedding 37 individually in paraffin (Kendall Paraplast X-tra, Tyco Healthcare, Mansfield, MA, USA) blocks. All paraffm-embedded samples were refrigerated (4 °C) during storage prior to use. Sectioning of Embedded Tissues Randomly, 4-5 paraffin-embedded hemi-maxilla samples were selected from each of the 8 experimental groups. In total, 34 (47%) specimens were selected from the group of 73 animals that comprised the complete set of available tissues harvested from the experiment. Two samples provided no useful sections for assessment. Starting from the palatal midline, tissues were cut in 8ji.m thick sagittal sections sequentially made towards the lateral aspect of the maxilla and floated on a distilled water bath heated to a consistent temperature of 36 °C. All sections that included molar teeth were mounted on 2% TESPA ((3-aminopropyl) triethoxysilane (Sigma-Aldrich, St. Louis, MO, USA) treated glass slides for staining. On average, molar teeth were visible on about 60 sections (average 4 sections/slide). Approximately every 4th slide was stained using hemotoxylin & eosin (Harris’s Alum Haematoxylin & Eosin Y). In addition, selected slides were prepared with using a modified Movat pentachrome staining technique (Schmidt, 1996). Quantification of Epitheial Migration All stained sections were assessed under light microscopy (Axioplan 2, Carl Zeiss Microlmaging GmbH, Munich, Gennany) with a lOx objective (Plan-NEOFLUOR lOxO,30) and images were captured with a digital charged-coupled device (CCD) objective (QICAM Mono 10 bit RGB 0.45x, Q Imaging, Surry, BC, Canada) and viewed 38 with a digital imaging program (Northern Eclipse, Empix Imaging Inc., Mississauga, ON, Canada) on a LCD computer monitor. Digital images extending from the interproximal region between the first molar (Ml) and second molar (M2) and between second and the third molar (M3) were taken and saved as Tagged Image File Format (TIFF) files with a calibrated measurement bar embedded into each image. Subsequently, the images were opened in ImageJ software (http:I/rsb.info.nih.gov/ij), calibrated for size and then the linear distance between the cemeto-enamel junction and the apical extent of the junctional epithelium was quantified for the distal root surface of Ml (MiD), the mesial root surface of M2 (M2M) and the distal root surface of M2 (M2D). M3 mesial root surface measurements of epithelial migration were not included in the data set due to the inconsistency with the presence of this tooth in all specimens and the exceptional variability in identifying acceptably oriented M3 tooth sections. The mesial surface of the first molar was not assessed due to the innate absence of premolars in mouse jaws. Anatomically, the junctional epithelium at the Ml mesial surface exhibits the characteristic orientation of the junctional epithelium with its apical loop (Page, 1982). This loop prevents accurate determination of the apical extent of the JE or PE as the apical loop of the gingiva may come into close approximation to the tooth surface. Figure 4 illustrates the method of quantifying epithelial migration at the distal surface of Ml and mesial surface of M2. All measurements were made by a single investigator (SA) and at a subsequent sitting a random subset of the group of images was re-measured with the same protocol outlined above, as a measure of intra-operator reliability. 39 Figure 4. Representative [_ section between iii c and second molar (lOx). CEJ (grey arrow heads). Apical extent of epithelial migration (white arrow heads). Length of epithelial migration measured for assessment (dashed line). Bar200 micrometers. Assessment of Periodontal Inflammation Digital images were taken from stained hemotoxylin & eosin stained sections viewed under a light microscope as described above. These standardized and calibrated images of the interproximal region between the first and second maxillary molars, were assessed for levels of inflammation associated with the periodontal tissues. Levels of inflammation associated with the inter-proximal gingival unit immediately subjacent to the JE or PE coronal to the alveolar bone, were scored according to the grading system described in table 1. 40 Score Histologic Description Mild cellular infiltrate with no appreciable 1 extension of rete ridges into the underlying connective tissue. Moderate cellular infiltrate with early 2 extension of rete ridges into underlying connective tissue. Severe cellular infiltrate with advanced 3 extension of rete ridges into underlying connective tissue. Table I. Qualitative assessment of levels of gingival Ieukocytic infiltrate as scored from histologic sections of the interproximal regions of the molar teeth. All samples were assessed individually at a single sitting. Assessment of the samples was preceded by a calibration exercise with representative images of the different levels of inflammation used as a reference. Figure 5 presents representative samples of gingival inflammation used for qualitative scoring of the gingival samples using the scoring system presented in table 1. 41 Figure 5. Representative H&E stained sections (I Ox magnification) of the inter-proximal gingiva and alveolar crest between the first and second molar used for qualitative scoring of gingival inflammation. Areas evaluated were immediately subjacent to the gingival epithelium and pocket epithelium. Bar= 250 micrometers. A. Score 1-Mild inflammation B. Score 2-Moderate inflammation C. Score 3-Severe inflammation. Assessment of Epithelial Migration into the Furcation Digital images were taken from stained hemotoxylin & eosin stained sections viewed under a light microscope as described above. Sections that were representative of the mid- furcation region of the first molar were identified as those sections that coincidently exhibited sections of the radicular pulpal tissues of the first molar. Sections of the molar teeth that failed to exhibit the radicular pulpal tissues were not considered as they were recognized to be sections further buccal or lingual to the mid-furcation region. Serial sections were assessed for the presence of epithelial tissue on the roof of the furcation. Each mouse sample was scored once in a dichotomous maimer for presence or absence of epithelial tissues in the mid furcation region based on assessment of all the available sections. Those samples, in which epithelial tissues were visible in the mid-furcation, 42 were scored positive for epithelial migration into the furcation. Figure 6 provides examples of intact furcation regions as well as sites that are representative of furcation epithelial invasion. Figure 6. Epithelial migration into furcation region as assessed on sagital sections of maxillary molar teeth. A) H&E staining of the mid bucco-lingual section of M I showing intact periodontal ligament and no evidence of epithelial migration. B) H&E staining of the bucco-lingual mid section of M I with evidence of epithelial migration (arrow). Bar=250pm. 2.2.3 Characterization of 136 mtegrm transgenic and gene knock-out mouse phenotypes by PCR Tail skin samples were utilized for characterization of the endogenous mouse 136 integrin and/or transgenic human 136 integrin gene expression for each strain of mice. Two millimeter frozen tail samples were ground to a fme powder in a stainless steal bowl cooled with liquid nitrogen until a uniform fine powder remained. Total ribonucleic acid (RNA) and genomic deoxyribonucleic acid (DNA) was isolated from the tissue powder using Aurum Total RNA Mini Kit (Bio-Rad, Hercules, CA, USA) according to the --- 43 manufactures instructions. For RT-PCR of human and mouse 6 integrin, one microgram of total mRNA was reverse transcribed using Oligo(dT) primer with the SuperScript First-Strand Synthesis System (Invitrogen Life Technologies, Carlsbad CA, USA), according to the manufactures protocols. For each PCR reaction, 2 il of reverse- transcribed RNA was combined with a total volume of 50 p1 buffer solution containing 20 mM Tris-HC1 (pH8.8), 10 mM KC1, 10 mM (NH4)2S04, 2 mM MgSO4, 0.1% Triton-X100, 10 mM dNTPs, 10 jM each of oligonucleotide primers and 1 jl (1 unit) of DNA polymerase. For the PCR amplifications, an initial denaturation step at 95°C for 2 minutes was followed by 45 cycles, consisting of denaturation at 94°C, annealing at 5 8°C, and extension at 72°C, each for 40 seconds. A fmal extension step was carried out at 72 °C for 5 minutes. The following primers were used for RT-PCR: Human 136 integrin: 5’- TCAGCGTGACTGTGAATATCC -3’ and 5’- GAGCACTCCATCTTCAGAGAC -3’ corresponding to 1266-11286 and 1786-1766 of human 136 eDNA (Accession no. NM_000888), respectively; mouse 136 integrin: 5’- AGGAGAATTTCACCCACCTG -3 and 5’- TGAATCTCTCGGCATCATCA -3 corresponding to 337-357 and 870-85 1 of mouse integrin 136 cDNA (Accession no. NM_021359), respectively. To identify the presence of the 136 integrin knock-out construct in the DNA of the mice, the primers for neoF-5’ CAGTAAATCGTTGTCAACAG and Km136R-5’ GTGGATCTGCTAAGTTAACC were used in the PCR reactions as described previously (Huang et al., 1996), The PCR products were separated by 1% agarose gel electrophoresis and stained with ethidium bromide for visualization under a UV light. 44 ct7 uwssossipspmsjososothnd iojpopujq-unSMipp1duwsoqi‘s4uomaInsouIouonbosqngssooodpuoI.u!.Iodxo oqnoqnonouiuoj.iomnuownsoqpoquosnoJosoidwsjjn‘Tu!pJooovuTn1s .io&udnoi2jnuowuodxooqopoinjonm‘ownujnoioumuonb!unnpouissnsn\.jnunun qontusnuqoowpopuqq-ouTsniopunouopOJOMuowoms’aow.iojposusojduinsiiv .I1HWX1JOBulpuiffi 2.4 All measurements were immediately recorded and subsequently transferred to an Microsoft Excel (Version 11.3 (Mac), Microsoft Corp., Redmond WA, USA) spread sheet. Data was compiled, confirmed and reviewed prior to direct transfer to SPSS (Version 16, SPSS Inc. Chicago, IL, USA) for further statistical analysis. Survival data was assessed for differences using Kaplan Meier Survival Analysis within the SPSS statistical package. Data sets were visually assessed for parametric distribution. Differences in dependant variables between the groups were assessed for statistical significance using ANOVA with a Bonferrom adjustment within the SPSS statistical package as well as using the online GrapbPad Software (http://www.graphpad.com/quickcalcs/ContMenu.cfln). An assessment of relative contributions of the independent variables to the observed differences between the groups was assessed using linear regression analysis within the SPSS software package. Statistical analysis of qualitative assessment of inflammation was performed using two- tailed Mann-Whitney U Test (VassarStats, http://faculty.vassar.edu/lowry/utest.html). Statistical analysis of epithelial migration into the molar furcations was performed using two-tailed Fisher’s exact test (GraphPad Software, http://www.graphpad.comlquickcalcs/contingency2.cfln). 46 CHAPTER 3: RESULTS 3.1 Survival Analysis Experimental diabetes induction with streptozotocin was predictable and maintainable over the experimental period in all groups. Diabetic animals showed overt signs of diabetes (polydipsia and polyuria) within 14-20 days subsequent to experimental induction. Seventy of the 112 animals completed the experimental protocol. Table 2 presents the numbers and gender of surviving animals. Gender differences were not apparent, consequently survival analysis was completed with male and female animals grouped together. Experimental WIT Cont WIT Exp 136-I- Cont 136-I- Exp 136 Rescue 136 Rescue K14136 K 14136Groups Cont Exp Cont # of Females 5/5 5/5 5/I I 3/I I 516 3/I I 5/5 4/8Surviving /Starting # #ofMales Surviving 4/4 6/6 7/9 3/10 316 214 5/5 5/6 IStarting # # of Animals Surviving 9/9 lI/Il (2/20 6/21 8/12 5/15 10/10 9/14 IStarting # (genders 100% 100% 60% 29% 67% 33% 100% 64% combined) Table 2. Presentation of numbers of animals surviving the I 6-week experimental period and total numbers of animals that began the experimental period. Kaplan-Meier survival curves for the experimental groups are presented in Figure 7. W/T control and experimental groups as well as the K14[36 control group showed no 47 premature death. The diabetic (36-I- showed the greatest rate of premature death with 71% loss during the experimental period. A significantly greater rate of premature death for the 136-I- group compared to the baseline FBV diabetic group was observed (P-value <0.05, ANOVA with a Bonferomn post hoc adjustment). The diabetic (36 rescue group showed elevated rates of premature death (67% loss) but did not reach statistical significance compared to the W/T diabetic group probably due to the fewer numbers of subjects in this group. H * 04 C) —WIT Control — W/T Diabetic ii — p6-I- Control — p6-I- Diabetic 02 6 Rescue Control — 6 Rescue Diabetic Kl46 Control — K14 B6 Diabetic 0.0- I I 0 5 10 15 20 Ex Weeks Figure 7. Graphical survival analysis for all groups during the I 6-week experimental period. Statistical analysis (ANOVA with a Bonferonni post hoc adjustment) revealed a significant difference for the experimental B6-I- group from both the WIT Diabetic and WIT Control groups. (*): P-value O.05. The remainder of the analysis will only consider the animals that completed the experimental period. All data from animals that died premature to the 16-week experimental period was removed for the subsequent analysis. 48 3.2 Pretreatment Wei2ht Assessment of pretreatment weight for the experimental and control groups within each of the 4 strains of mice revealed no significant difference (data not shown) (Student t-test: P-value >0.05). Experimental and control groups for each strain were combined and analysis revealed a significant difference between the groups (P-value <0.0001, ANOVA with a Bonferonni post hoc adjustment). Further analysis revealed the 6-/- group (8 females, 10 males) (29.9 g ± 4.2) to be significantly lighter than the W/T strain (10 females, 10 males, 38.0 g ± 6.1, P-value <0.01), the 36 rescue (8 females, 7 males, 36.1 g ± 6.1, P-value <0.01) and the K146 group (9 females, 10 males, 38.0 g ±8.4, P-value < 0.00 1). Figure 8 presents the data pertaining to pretreatment weight. Pretreatment Weight .c • W/T • B6-/- B6 Rescue • K14/86 Figure 8. Assessment of pretreatment weight for different strains ± one standard deviation. Animals were weighed prior to any interventions. Statistical analysis was performed using ANOVA with a Bonferonni post hoc adjustment. ): P-value <0.05, (©): P-value<0.0 I, (): P-value 0.001 49 33 Experimental Diabetes All strains of mice were assessed for levels of glycemic control. Control groups maintained blood glucose levels consistent with non-treated mice throughout the experimental period (Lalla et al., 1998a; Like and Rossini, 1976). Experimental groups all developed hyperglycemia of greater than or equal to 306 mgldL (17 mmolfL) as measured by fasting tail blood samples, which is consistent with a severely diabetic state (Lalla et al., 1 998a). Statistical testing for differences between the levels of hyperglycemia within the strains as well as within the experimental and control groups was performed using one-way ANOVA with a Bonferonni post hoc adjustment. All experimental groups achieved highly significant levels of clinical hyperglycemia compared to control groups of the same strain (P-value <0.0000 1). No significant differences for fasting blood glucose levels during the experimental period were revealed amongst the experimental groups (P-value> 0.05). Similariy, amongst the control groups no significant differences were noted for fasting blood glucose levels during the experimental period (P-value >0.05). Figure 9 displays fasting glucose blood levels for all groups and statistical significance. 50 •W/T Cont (n=9) WIT Ex (n=11) B6-/- Cont (n=12) B6-/- Ex (n=6) 66 Rescue Cont (n=9) B6 Rescue Ex (n=6) K14B6 cont (n=1O) K14B6 Ex (n=1O) Figure 9. Fasting blood glucose of all animals during experimental period as assessed with tail blood samples ± standard deviations. Differences between groups was determined using ANOVA with a Bonferonni post hoc adjustment. No differences were noted among the different diabetic groups. Similarly, no significant differences were noted among the control groups. (*): P-value <0.05, (‘): P-value<0.0 I, (‘): P-value <0.001, (**): P-value <0.0001 40 Fasting Blood Glucose Levels J30 0 E E 20 34 Enithelial MH!ration Epithelial migration was assessed from 8 jim thick sagitally sectioned, stained and mounted paraffin-embedded maxilla. All dimensions were quantified on digital images using ImageJ software. The sagittal orientation of the sections allowed interproximal assessment of MiD, M2M and M2D epithelial migration. Table 3 presents data for epithelial migration from the 8 experimental groups. In total, 92 separate surfaces were measured. All surfaces were measured twice at separate sittings, blinded to the experimental group as well as any previous measurements. Replicate measurements from each surface were averaged to provide a single measure for that surface which was used in all subsequent analysis. 51 Experimental WIT Cant WIT Exp p6-I- Cont p6-I- 6 Rescue 6 Rescue KI46 KI46 ExpCont ContGroups (n14) (n11) (nI2) (n12) (n=12) EXP(n=9) (n=12) (i’iIO) Epithelial 47.42iim± 57.27pm± 199.l8jim 235.75jm 143.95pm 131.94pm 75.O8pm± 63.89pm± Migration I 1.68 21.78 ± 60.93 ± 134.06 ± 73.35 ± 44.40 43.64 47.99 Table 3. Epithelial migration as measured at the M I distal, M2 mesial and M2 distal tooth surfaces. Measurements performed on digital images taken of paraffin sections and measured using lmageJ software. Number for each group represents numbers of surfaces measured in each experimental group. Statistical analysis for differences between the groups was performed using one-way ANOVA with a Bonferonni post hoc adjustment. No significant differences were identified between the experimental and control groups within each strain. Figure 10 illustrates the results of the epithelial migration at MiD, M2M and M2D tooth surfaces for each of the 8 groups. Insufficient data prevented gender analysis. 52 Epithelial Migration 450 •W/TCont(n=14) 400 ___________ I W/TExp(n=11) 350 I B6-/- Cont(n=12) 300 250 E:p(n=12) ‘ 200 B6 Rescue Cont(n 12) z T 150 I B6 Rescue(n=9) iii • K14B6 Cont(n=12) Ii •j K14B6 Exp(n=10) Figure I 0. Epithelial migration as measured at the M I distal, M2 mesial and M2 distal tooth surfaces. Measurements performed on paraffin sections measured using lmageJ software. Number for each group represents numbers of surfaces measured in each experimental group. Statistical differences between groups were identified using ANOVA with a Bonferonni post hoc adjustment (*): P-value <0.05, (“): P-value<0.0 I, (©c): P-value <0.001 Data analysis reveals that control W/T mice consistently exhibit the least amount of epithelial migration regardless of experimental or control treatment, with 57.27 jim ± 21.78 and 47.42 p.m ± 11.68 respectively. Control f36-/- mice exhibit significantly greater epithelial migration (P-values <0.001) compared to the W/T control mice, with 199.18 p.m ± 60.93 and 47.42 p.m ± 11.68 of epithelial migration noted respectively. Epithelial migration observed in the control K146 group was significantly lower than the control 136-/- group (P-value <0.001) with 75.08 p.m ± 43.64 and 199.18 p.m ± 60.93 of epithelial migration noted respectively. No significant differences (P-value >0.05) in epithelial migration was noted between the control W/T and control K146 groups, with average epithelial migration noted at 47.42 p.m ±11.68 and 75.08 p.m ± 43.64, respectively. Epithelial migration noted in the control fE6 Rescue group (143.95 p.m ± 73.35) was 53 significantly higher than the control W/T group (P-Value<O.O1), reaching an intermediate level between the W/T and 136 knock-out groups. Since no statistical differences were seen between the experimental and control groups for each strain, they were combined to facilitate analysis between the different strains. Figure 11 presents the graph of epithelial migration for each strain. As is apparent, once the experimental and control groups were combined, there was a highly statistical difference between all groups except the K14[36 and WIT groups where no difference was noted. Epithelial Migration 400 350 _____________ •W/T(n=25) E 0 250 B6-/- (n=24) 2200 2 150 B6 Rescue(n=21) .!ioo a) K14B6 (n=22) 0 _.J* Figure I I. Epithelial migration as measured at the M I distal, M2 mesial and M2 distal tooth surfaces ± standard deviation. Measurements performed on paraffin sections measured using lmagej software. Number for each group represents numbers of surfaces measured. Statistical analysis done using ANOVA with a Boriferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.0 I, (): P-value <0.001, (): P-value<0.000I 3.5 Epithelial Migration into the Furcation As a further assessment of epithelial migration, Ml and M2 furcation regions were evaluated histologically for presence of epithelium in the furcation area. Furcation regions were dichotomously graded for the presence or absence of epithelium in the mid- 54 III furcation region. In total, 61 Ml and M2 furcation regions were available for assessment in 32 sectioned maxilla. Data was presented as an aggregate of epithelial migration at both the Ml and M2 sites. Table 4 presents the data for epithelial invasion into the furcation. Experimental Groups (36 Rescue 36 Rescue K 14136 K 14136WIT Cont WIT Exp (36-I- Cont (36-/- Cont Exp Cont Exp(n: number of (n=8) (n=7) (n=8) (n=8) molars (n=8) (n=6) (n8) (n8) evaluated) Epithelial Migration into 0% 0% 37.5% 62.5% 37.5% 33.3% 0% 12.5% Furcation Table 4. Proportion of sites in each experimental group, for which epithelium was noted in the M I and M2 furcation regions as determined histologically on maxillary sections. Data is an aggregate of the M I and M2 sites. “n” represents the total number of furcation areas evaluated in each group. Statistical testing was performed using two-tailed Fisher’s exact test. Greater proportions of epithelial furcation invasion were noted in the 36-/- group and the 6 rescue group, though no statistical differences were identified. As no statistical differences were noted between the control and experimental groups in each strain the experimental and control groups for each strain were combined to allow statistical testing between mouse strains. The 6-/- strain and [6 Rescue strain showed statistically greater proportions of furcation areas with epithelial invasion compared to the W/T group, with P-values of 0.002 and 55 0.017, respectively. Similarly, the proportion of epithelial migration into the furcation for the 136-I- was higher than that seen in the K14f36 group, reaching significance (P-value =0.015) while the difference between the K14136 and [36 Rescue group approached statistical significance (P-value 0.07). No statistical differences were noted between the W/T and K14[36 groups. Figure 12 presents data for epithelial migration into the molar furcation region. Epithelial Migration into Furcation 75% I IC 50% I I 25% 0% — -- WIT (n=15) B6-/- (n=16) B6 Rescue (n=14) K14B6 (n=16) Figure 12. Proportion of M I and M2 furcation regions that exhibited epithelial tissues in the furcation as measured histologically in the maxilla. Numbers of molar teeth assessed in each group is presented. Statistical differences between groups were identified using two-tailed Fisher’s exact test. (#): P-value 0.07, (*): P-value <0.05, (n): P-value<0.0 I, (°): P-value <0.001 3.6 Alveolar Bone Loss As an assessment of alveolar bone loss, de-fleshed and stained mandibular molar teeth were assessed for area of root exposure between the alveolar crest and CEJ on the lingual surface. From the total group of 72 animals, a random subset of 26 samples was blindly re-measured as an assessment of reliability (Pearson correlation, two-tailed). 56 Measurements of the first molar showed statistically significant correlation of 0.942 at a P-value of 0.01. Repeated measurements of the second molar showed increased variability and a non-significant correlation of 0.80 at a P-value of 0.01. Since measurement of first molar teeth was highly reliable further analysis was restricted only to the first molar data. Variations in tooth size may affect measurements of exposed root surface area. Owing to the significantly reduced weight seen in pre-treatment weight of 36-/- mouse strain and the reported positive or negative correlations between body mass and tooth size (Carranza and Perez-Barberia, 2007; Kavanagh et al., 2007), mandibular first molar tooth width was evaluated as a measure of tooth size. Table 5 presents data on Ml tooth width from each experimental group as well as gender aggregates within each group. 57 Experimental WIT Cont WIT Exp 136-I- Cont 136-I- Ex1 136 Rescue 136 Rescue K 14136 K 14136Groups (n10) (n10) (n10 (n=6) Cont Exp Cont Exp(n-female) (n8) (nl2) (n14) (n6) (n=l0) (n=6) (n10) (n8)(n-male) (n6) (n=8) (n= 10) (n= 10) First Molar . l.4Omm± l.4lmm± l.44mm± I.38mm± l.44mm± l.45mm± l.4Omm± 1.38mmWidth- 0.03 0.04 0.05 0.04 0.03 0.02 0.03 ±0.04Female First Molar l.4lmm± l.43mm± l.43mm± l.46mm± I.47mm± l.48mm± l.4Omm± 1.41mmWidth 0.04 0.02 0.04 0.01 0.01 0.02 0.03 ±0.02Male First Molar I.4Omm± l.42mm± I.43mm± l.42mm± l.46mm± I.47mm± l.4Omm± 1.39mmWidth- 0.03 0.03 0.05 0.05 0.03 0.02 0.03 ±0.03Combined Table 5. First molar mesio-distal tooth width as measured from cle-fleshed and stained mandibular jaws digitally imaged under 40x magnification and measured using ImageJ software. Statistical analysis was performed using one-way ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (°): P-value<0.O I, (‘): P-value <0.001 Generally gender was not found to affect differences in tooth size except for a significant difference associated with the female experimental f36-I- group exhibiting significantly reduced molar crown mesio-distal width compared to the male experimental 6-/- group (1.38 mm ± 0.04 and 1.46 mm ± 0.01 respectively). Owing to only a few differences of mesio-distal crown width being ascribed to gender, statistical analysis for differences in tooth size between the experimental groups was performed on aggregates of female and male samples using one-way ANOVA with a Bonferonni post hoc adjustment. Results revealed no differences between the experimental and control groups within each strain of mice. The same analysis showed no statistical differences between the W/T groups and [6-/- and between the W/T and f6 Rescue groups with respect to tooth size. However, the K146 groups exhibited significantly smaller Ml molar width compared to 58 the 6 Rescue groups. Figure 13 presents data of the first molar widths for each of the experimental groups. Molar Tooth Width 1.55 • WIT Cont (n=9) 1.50 W/T Exp (n=11) E E B6-/ Cont 1.45 B6-/- Exp 1.40 (n=6) B6 Rescue Cont (n=8) 1.35 B6 Rescue Ex (n=7) X Kl486Cont 1.30 (n=1O) K14B6 Exp -, (n=9) 1.25 Figure 13. First molar mesio-distal tooth width as measured from de-fleshed and stained mandibular jaws digitally imaged under 40x magnification and measured using lmageJ software. Statistical differences between groups were identified using ANOVA with a Bonferonni post hoc adjustment. (*.): P-value <0.05, (**): P-value<0.Ol, (*): P-value =0.001 Bilaterally, de-fleshed and stained mandibular jaws were assessed for alveolar bone loss represented by root exposure between the CEJ and alveolar crest as measured on digital images of the stained lingual surfaces and quantified by Image J software. Right side and left side Ml teeth were assessed independently of each other and results were aggregated in the final data set. Table 6 presents the data regarding alveolar bone loss from the first molar of all groups for both male and female mice and a combined average of the entire group. 59 Experimental 6-/- Cont 6 Rescue 6 Rescue K I 46 K I 46 ExpW/T Cont WIT Exp (n= 10 6-/- E Cont ContGroups (n10) (nI0) (nl4) (n=6) (nl0) Exp (n=6) (n=8)(n-female) (=8) (n12) * (n6) (n=8) (n10) (10)(n-male) (n6) (n= 10) First Molar 0.44mm2 0.40mm2 0.49mm2 0.44mm2 0.52mm2 0.45mm2 0.42mm2 0.39mm2Bone Loss ± 0.04 ± 0.03 ± 0.05 ± 0.05 ± 0.05 ± 0.06 ± 0.06 ±0.02Female First Molar 0.37mm2 0.35mm2 0.39mm2 0.46mm2 0.49mm2 0.52mm2 0.42mm2 0.44mm2Bone Loss ± 0.02 ± 0.07 ± 0.05 ± 0.06 ± 0.07 ± 0.21 ± 0.06 ±0.05Male First Molar 0.41mm2 0.37mm2 0.43mm2 0.45mm2 0.51mm2 0.49mm2 0.42mm2 0.42mm2Bone Loss ± 0.05 ± 0.06 ± 0.07 ± 0.04 ± 0.06 ± 0.16 ± 0.06 ±0.04Combined Table 6. First molar bone loss quantified as the area between the CEJ and alveolar crest, assessed from de-fleshed and stained mandibular jaws digitally imaged under 40x magnification and measured using ImageJ software. Statistical analysis for gender differences performed using one-way ANOVA with a Bonferonni post hoc adjustment. (*): P-value <0.05, (**): P-value<0.0 I, (): P-value <0.001 3.6.1 Combined Male and Female Bone Loss To assess if gender played any role in the observed differences, statistical analysis using one-way ANOVA with a Bonferonni adjustment was applied to the data set and revealed no significant differences between gender except in the control 136-I- group where the female 136-I- control group exhibited greater bone loss than the male 136-I- group (P-value <0.05), with 0.49 mm2± 0.05 and 0.39 m2± 0.05 bone loss noted, respectively. In light of only modest differences that could be ascribed to gender, analysis of the alveolar bone loss data was completed with both genders combined, with statistical analysis completed using a one-way ANOVA with a Bonferonnoi post hoc adjustment. Figure 14 represents graphically the results for alveolar bone loss for each of the experimental groups. 60 Alveolar Bone Loss 0.8 0.6 E 41 U u 0.4 0 0 0 0. 0.2 RK14/B6Corit (n=20) K14/B6 Ex (n=18) 0.0 Figure 14. Quantification of surface area between the CEJ and alveolar crest of the first molar ± standard deviation as measured by lmageJ software. Statistical analysis for differences between groups supplied byANOVA with a Bonferonni post hoc adjustment. No differences were noted between control and experimental groups within each strain. Number for each group represents numbers of teeth examined. (*): P-value <0.05, (u): P-value <0.01, (): P-value 0.00 I Data analysis revealed no differences between the experimental and control groups within each strain of mice. Analysis of relationships between the different strains revealed that bone loss was greatest in the 136 Rescue strain, with the control 136 Rescue strain achieving statistically significant greater bone loss compared to the control W/T group (0.51 mm2± 0.06 and 0.41 mm2± 0.05, respectively, P-value <0.01), compared to the control 136-I- group (0.51 mm2± 0.06 and 0.43mm2± 0.07, respectively, P-value <0.05) and compared to the K14136 group (0.51 112 ± 0.06 and 0.42 2± 0.06, respectively, P-value <0.01). Differences between the experimental K14f36 group and the others experimental groups was not achieved, possibly due to the markedly increased variance associated with the experimental 136 Rescue bone loss measurements. 61 E E U 0 0 0 x UI • WIT Cant (n=10) WIT Ex (n=10) B6-/- Cant (n= 10) B6-/- Lx (n=6) 56 Rescue Cant (n= 10) B6 Rescue Ex (n=6) 3.6.2 Alveolar Bone Loss - Female While differences between genders were not generally apparent as discussed above, when data for bone loss was assessed in a gender specific manner for to allow comparison with previous studies (see discussion below). The resulting analysis revealed different trends between male and female mice. Female diabetic groups showed a consistent trend towards less alveolar bone loss compared to the non-diabetic control group of the same strain. Figures 15 and 16 display the graphical results for alveolar bone loss for the female and male groups, respectively. First Molar Alveolar Bone Loss - Female Figure 15. Quantification of surface area between the CEJ and alveolar crest of the first molar alveolar bone loss ± standard deviation for female mice as measured from digital images by lmagej software. Statistical analysis provided by pair-wise comparisons using Student t-test. (#): P-value 0.05-0.08, (*): P-value <0.05, (): P-value =0.001 Comparison for the female experimental and control groups within each strain revealed significantly less bone loss (P-value =0.042, Student t-test) in the experimental W/T 62 verses the control W/T female group, with 0.40 mm2 ± 0.05 and 0.44 mm2± 0.04 of bone loss noted, respectively. A similar comparison of the experimental 136 Rescue and control 136 Rescue groups revealed significantly less alveolar bone loss in the former group (P-value = 0.02 1, Student t-test), with 0.45 mm2± 0.06 and 0.52 mm2 ± 0.05 of bone loss noted, respectively. Analysis of the alveolar bone loss for the experimental f36- I- group and the control 136-I- group showed a similar trend with less bone loss in the former with 0.44 mm2 ± 0.05 and 0.49 ± 0.05 of bone loss noted, respectively. However, the difference did not reach statistical significance (P-value =0.079, Student t test), probably due to high mortality in these groups resulting in fewer numbers of teeth available for assessment. 3.6.3 Alveolar Bone Loss - Male Assessment of alveolar bone loss between the male experimental and control groups within each strain revealed that the induction of experimental diabetes increased the levels of bone loss that the experimental groups experienced, in direct contrast to the observations in the female group comparisons. The 13 6-I- group analysis revealed a statistically greater level of bone loss in the male experimental 136-I- group compared to the male control f36-/- group (P-value=0.0l, Student t-test) with 0.46 mm2± 0.06 and 0.39 mm2± 0.05 of bone loss noted, respectively. Differences between the male experimental 136 Rescue group and the male 136 Rescue control group as well as between the male experimental K14f36 group and the male control K14136 group showed trends to greater bone loss for the experimental groups over the control groups but these differences did not reach significance. No statistical differences were noted between the 63 Figure 16. Quantification of surface area between the CEJ and alveolar crest of the first molar alveolar bone loss ± standard deviation for female mice as measured from digital images by lmagej software. Statistical analysis provided by pair-wise comparisons using Student t-test. (*): P-value <0.05, (‘): P-value <0.001, (): P-value =0.001 male W/T experimental group and the male WIT control group with 0.35 mm2 ± 0.07 and 0.37 im2± 0.02 of bone loss noted, respectively. First Molar Alveolar Bone Loss-Males F •W/TCont (n=8) L I’ r LiiI 0.1 0.0 B6 -I- Cont (n=14) B6-/- Ex (n=6) B6 Rescue Cont (n=6) B6 Rescue Ex (n4) K14B6 Cont (n=10) K14B6 Ex (n= 10) 3.7 Gin2ival Inflammation Levels of inflammation immediately subjacent to the gingival epithelial tissues were graded histologically at the interproximal sites between Ml & M2 and between M2 & M3. In total 59 different sites were evaluated from 30 specimens. The adjacent interproximal sites were graded in a blinded manner to avoid measurement bias. The results for each experimental group were aggregated for statistical analysis and graphical representation. Kappa assessment, as a measure of intra-examiner agreement of blinded replicate measurements revealed a significant level of agreement between scores (Kappa =0.68, P-value<0.00 1), suggesting good intra-examiner agreement (Altman, 1991). The 64 0% WIT WIT Ex B6-/- 86-I- Ex 86 86 K1486 K1436 Corit (n=8) Cont (n=8) Rescue Rescue Cont Ex (n7) (n8) (n=8) Cont Ex (n=6) (n=6) (n=8) Figure 17. Qualitative assessment of inflammation immediately subjacent to the JE and/or PE as measured histologically from from 30 animals. 30 interproximal sites evaluated at the M l/M2 region and 29 interproximal sites evaluated at the M21M3 region. Fesults from both the M I /M2 and M21M3 regions aggregated together. Statistical analysis performed using Mann-Whitney U test (pair-wise assessments). (): P-value<0.0 I, (): P-value =0.001 Assessment of differences between the control groups reveled that compared to the control W/T group, the control f36-/- group and the control 36 Rescue group both exhibited significantly greater levels of gingival inflammation, with P-value =0.0009 and P-value =0.004 respectively (Mann-Whitney U test). No statistical difference was found for graded levels of inflammation between the control K14136 group and control W/T group or between the control and experimental groups for all mouse strains. In a fashion similar to the assessment of epithelial invasion into the furcation and epithelial migration, Kappa measurement result suggests that the level of inflammation assessed was reproducible and relatively reliable for the sample assessed. Figure 17 represents the cumulative histogram of aggregate scores of inflammation. Periodontal Inflammation-Experimental Groups 100% I 75°h 50% 25% 0. • Severe 9Moderate lMiId 65 the data from the control and experimental groups were aggregated. Figure 18 shows the proportions of gingival inflammation scores for the different mouse strains. Periodontal Inflammation-Mouse Strains 100°k 75% • Severe Moderate 50% lMiId 0 25% 0% FBV (n=16) B6-/- (n=16) B6 Over (n=13) Figure 18. Qualitative assessment of inflammation immediately subjacent to the JE and/or PE as measured histologically from sagitally sectioned maxilla from 30 animals. Experimental and control groups for each strain were combined into mouse strain groups. (30 interproximal sites evaluated at M I & M2, 29 interproximal sites evaluated at M2 & M3) Statistical analysis performed using Mann-Whitney U test (pair-wise assessments). No statistical difference was noted between the WIT strain and the K 1486 strain. (#): P-value =0.06, (*): P-value <0.05, (‘h’): P-value<0.0 I, (‘): P-value<0.O0 I The 136-!- strain and the 136 Rescue strains exhibited the greatest proportion of sites with severe inflammation, not being significantly different than each other. In addition, to these two strains exhibiting statistically greater inflammation scores than the W/T group, the 136-I- and strain showed significantly greater inflammation scores than the K14136 strain (P-valueO.02, Mann-Whitney U test), while the difference between the 136 Rescue strain and the K14136 strain approached significance (P-value=O.06, Mann-Whitney U test). III B6 Resc (n=14) 66 3.8 Characterization of (36 Inteirin Expression in the Transenic and Knock-out Mouse Lines PCR results from the three (36 integrin transgenic and knock-out mouse lines used in the present study is presented in Figure 19. Visualization of PCR and RT-PCR products using gel electrophoresis confirm the absence of murine (36 integrin in the (36-I- strain, while the K14(36 strain exhibits the human 136 integrin gene in addition to the murine (36 integrin gene and the (36 Rescue strain expresses only the human (36 integrin gene. 136 -I- K14(36(36 Rescue Wild-type ) It) & E E E • It) C ‘ a) C C ‘ — LL - E E - 0 0 0 0 — c E .c E . c E .c F . human j36 mouse 136 — neoR5’ & Km13R-5’ — Figure 19. Characterization of the 6 integrin transgenic and knock-out mouse lines by PCR and RT-PCR.The PCR products were separated by electrophoresis in a I % agarose gel, stained with ethidium bromide and visualized under a IJV light. Results confirm that (36-I- strain is deficient in murine and human 136 integrin genes while K14 (36 strain express both murine and human (36 integrin genes. The (36 Rescue strain expresses only the human (36 integrin gene while the wild-type mice express the mouse 6 integrin only. 67 CHAPTER 4: DISCUSSION This study’s results concur with published reports that deficiency of integrin 136 is associated with elevated levels of chronic inflammation both at extraoral epithelial tissues (Huang et al., 1996; Sheppard, 2001) and with the more recently published novel findings that the absence of integrin 136 is associated with periodontal disease initiation and progression (Ghannad et a!., 2008). Furthermore, our findings that the re-introduction of the human 136 integrin sub-unit in the 136 Rescue mouse strain and it’s over-expression in the 136 over-expressing mouse strain, to varying degrees, reduce the level of gingival epithelial migration, compared to 136 knock-out mouse strain, reinforce the notion that the constituent expression of integrin av136 within the junctional epithelium is directly associated with periodontal stability. In otherwise healthy individuals periodontal disease, as mediated by dental accumulations of bacterial plaque, is characterized by local inflammation that unpredictably results in localized areas of periodontal attachment loss (Haffajee and Socransky, 1986; Lindhe et al., 1983; Loesche and Grossman, 2001; Page and Schroeder, 1976; Tu et al., 2004). Much of the recent attention towards the pathogenesis of periodontal disease has turned to factors that alter the inflammatory process since it is the local inflammatory process that ultimately leads to the majority of the attachment loss (Chapple, 2009). Recently published observations that the absence of integrin av136 is associated with increased periodontal inflammation, epithelial migration and exaggerated 68 periodontal bone loss, underpin the central role local inflammatory response plays in periodontal disease (Ghannad et al., 2008). Central to the purported mechanism between integrin av136 and periodontal stability is the activation of latent TGFI31 by integrin av6. The epithelial-restricted integrin av136 is as been shown to be an extremely effective ligand for binding to the RGD tri-peptide sequence of TGFI31-LAP. When a traction force is applied by the epithelial cell on the ECM bound TGFI3 LLC, a conformation change will be induced in the TGFI3 LLC, exposing the TGFI31 binding site for its receptor (Munger et al., 1999; Sheppard, 2006). Once exposed, the TGF3l binding site will bind to the membrane associated TGFI31 receptors (Tt3R) and initiate the subsequent intracellular cascade mediated by SMAD proteins (Itoh and ten Dijke, 2007). TGF(31 is well recognized as having a potent dampening effect on local inflammation (Kulkarni et al., 1993). As the JE is chronically associated with inflammation, even under experimental protocols of meticulous plaque control (Page, 1982), it might seem intuitive that the dento-gingival complex may benefit from localized TGFf31 activation to reduce the potential collateral damage caused by this chronic inflammation. The observation that increased levels of cytoplasmic messenger RNA for TGFI31 in the gingival tissues is associated with periodontal health (Dutzan et al., 2009) is consistent with the observation that JR constitutively expresses integrin avf36 (Ghannad et al., 2008), a potent local activator of TGFI31, but more importantly provides a potent local mechanism for TGF1 activation. 69 As it is conceivable that the f36 knock-out strain of mice could be predisposed to periodontal disease by some as yet undiscovered gene inadvertently inactivated while developing the 136 -7- strain (Sheppard, 2001) as opposed to the loss of the 136 integrin, we sought to test weather a transgenic 136 integrin over-expressing strain and a 136 Rescue strain would effectively compensate for the loss of the murine endogenous 136 integrin. If these murine strains could exhibit a measure of protection from periodontal disease compared to the 136 knock-out strain, then the evidence supporting local periodontal protection by integrin av136-activation of TGF131 may be further strengthened. Although a central feature of periodontal disease is clinical attachment loss (Armitage, 1999), it is necessarily preceded by the apical migration of the JE. The process of apical migration involves proliferation of the JE cells and its migration into the underlying connective tissues, both of which are classically associated with the altered pocket epithelium and histologic signs of periodontal disease (Muller-Glauser and Schroeder, 1982). TGF(31 is well known to arrest cellular growth in many cell types including epithelial cells (Sheppard, 2005). The mechanisms associate with attenuation of epithelial proliferation is believed to stem from Smad and FoxO transcription factor up- regulation of cyclin-dependant kinase inhibitors 21 Cipi/WAF1 and ,15 Ink4b (Gomis et al., 2006). Additional TGF13 1-mediated cellular mechanisms related to this repression of epithelial cell growth are related to transcriptional repression of c-Mye, a pro-growth transcription factor, and the inactivation ofp70 S6K, a serine/threonine kinase promoter of ribosome biogenesis and cell growth (Rahimi and Leof, 2007). Overall, TFG(31 70 activation within the dento-gingival complex would serve to reduce epitheial proliferation, potentially offering a significant contribution to stabilization of the JE. This hypothesis is consistent with high levels of TGFf31 gene activity observed in clinically healthy ginginval tissues (Dutzan et al., 2009). Furthermore, the observation that PE is deficient in expression of integrin av136 (Ghannad et al., 2008) and therefore associated with potentially reduced activation of TGFI31, is consistent with the migratory phenotype of PE as it invades the underlying CT (Muller-Glauser and Schroeder, 1982) and the clinical observation of significantly reduced levels of TGFI31 gene expression in gingival tissues associated with periodontal disease (Dutzan et a!., 2009). Our observations revealed that the 136-I- mouse strain developed approximately 4x greater apical migration of the JE compared to WIT groups. These findings are consistent with recently published observations (Ghannad et al., 2008). Furthermore, our results show that the control K 14136 strain of mice exhibited levels of epithelial migration similar to the control W/T strain of mice, with no statistical differences noted between the two groups. However, it may be suggested that since these mice express the normal murine gene for the integrin 136 subunit in addition to the human 136 integrin transgenic gene, the observed effect could be attributed solely to the murine 136 integrin gene. Of importance, however, is that the resulting human integrin 136 transgenic gene does not interfere with the normal tissue homeostasis and predispose the strain to periodontal disease progression. Our observations revealed 136 Rescue strain of mice show intermediate levels of epithelial migration between the control WIT strain and the control 136-I- strain. Statistical analysis revealed less epithelial migration from the (36-I- group (P-value 71 <0.00 1, ANOVA with Bonferonni post hoc adjustment), suggesting that 136 reintroduction under the control of K14 partially rescued the strain from this purported feature of murine periodontal disease. Knowing cytokeratin 14 is identifiable in healthy JE basal cells and the PE during disease suggests that h136 integrin gene expression under the K14 gene promoter is likely. However, the tempro-spatial expression of h136 integrin protein and its ability to co-localize with the av integrin subunit as a functional membrane integrated integrin heterodimer in the JE is to date, unknown. It is possible that defects in formation of the transgenic heterodimer or reduced levels of its activity prevented a complete periodontal rescue, however, the observation of a partial rescue for epithelial migration further supports the hypothesis that integrin av136 activation of TGF131 locally confers a measure of periodontal protection in this murine model. Previous studies assessing natural progression of periodontal disease in isolated mouse groups identified furcation bone loss as a good measure of disease progression (Wiebe et al., 2001). In contrast to assessing furcation invasions directly from mandibular jaws as they reported, we histologically assessed the furcation region with the opinion that epithelial migration into the furcation may be a more sensitive measure of early changes, particularly considering we only had a 16-week period of experimental diabetes. The greatest proportion of sites exhibiting furcation epithelial invasion was noted in the 136-/ strain, consistent with the results presented above, for epithelial migration. The K14136 strain exhibited levels of epithelial migration statistically similar to the W/T strain while. The 136 Rescue strain, though expressing a marginally increased proportion of furcation 72 sites with intact furcation regions, was not statistically significant from the f36-/- strain (Fisher exact test, two tailed, P-value >0.05). For clinicians, the most relevant feature of periodontal disease is periodontal pocket development and attachment loss that ensues following alveolar bone loss. Practically, this is of critical importance in the human condition where progressive loss of alveolar bone and clinical attachment will result in tooth loss. Although TGFI31 is not directly associated with bone formation (Sampath et al., 1987), it is indirectly associated with bone stability through its local anti-inflammatory effects associated with suppression of the cell and humoral immune response (Gurkan et al., 2005). Reduced inflammatory stress encourages bone stability by increasing the osteoprotegerin (OPG): receptor activator of nuclear factor KB ligand (RANKL) ratio, and thereby opposes the formation of osteoclasts (Katagiri and Takahashi, 2002; Taubman et a!., 2005). Furthermore, TGFI31 has been directly associated with stability of the ECM by reducing matrix metalloproteinase (MMP) production and increasing cellular production of tissue inhibitors of MMP’s (TIMP5), which can significantly enhance the breakdown of collagens, a primary component of the ECM (Overall et al., 1991). Our measurements for exposed root surface for the control WIT mice are in agreement with those recently published (Ghannad, 2007), with average bone loss reported in their results of 0.40mm2±0.055 agreeing very well with 0.41 mm2± 0.05 of bone loss reported by our results. However, while measurements of alveolar bone levels of the control 136-/ strain (0.43 mm2± 0.07) found in the present study were not significantly different from 73 the control WIT group, Ghannad et.al. (2008) reported significant differences between the 136-I- group (0.50 im2± 0.114) and the W/T group, in their 12 -month old animals, the most comparable group to the groups used in the sample presented here. The differences may be related to the gender composition of the experimental groups which was heavily biased to female animals, with 9 female/5male and 17 female/i male animals used in the 12 month old 136-I- and W/T groups, respectively (Ghannad, 2007). Surprisingly, though our initial analysis showed only limited gender differences, when we compared the results of our female sample only with these previously published results (Ghannad, 2007), our results were in agreement, with 0.50 in2± 0.114 of bone loss reported for the 12 month 136-I- group reported (Ghannad, 2007) compared to 0.49 mm2± 0.05 measured for our 136-I- female mice. In a similar fashion to the statistical analysis used by that study, paired Student t-tests we were able to corroborate a statistically significant increase (P-value <0.05) in bone loss of the female 136-I- over the female members of our control W/T group (0.44 mm2± 0.04). An additional confounding factor that may not permit direct comparison of our results to those previously published by Ghannad et.al. (2008) is related to the older age of our sample at sacrifice. With an average age at sacrifice of 14.4 months ± 1.04, our animals are 20% older than the closest age group reported by Ghannad et.al. (2008). As periodontal attachment loss is accepted to accumulate over a life time (Loe et al., 1986) with no natural means of regaining lost attachment, the differences between the control and experimental groups may be diminished in older subjects. As our animals were older 74 at sacrifice we may expect that a diminished difference may exist between the WIT and 136-I- groups. The lack of a significant difference between control 136-I- group and the control W/T group in terms of measuring alveolar bone loss seems in contrast to the results of epithelial migration reported above but may be explained by the technique employed for measuring bone loss. The approach used for measuring alveolar bone loss reveals the entire area of root exposure not only the difference implicated to the experimental treatment used. However, this area of root exposure includes the initial area of the root exposed by physiologic requirements for health (Garguilo, 1961), in addition to any further root area exposed by horizontal bone loss. The physiologic area of exposed root surface required for the normal development and maintenance of a healthy dento-gingival physiologic structure is referred to as biologic width in humans (Ingber et al., 1977). In the murine model of periodontal disease, the area of exposed root surface which corresponds to the development of a normal dento-gingival apparatus may be quantified by the area of exposed root surface measured in young W/T mice, and may be referred to for the purposes of this discussion as physiologic root exposure (PRE). In mice as early as 6-12 weeks of age, PRE is readily apparent (Ghannad et al., 2008; Lalla et al., 1998a). Relative to the area of the area of root exposure that develops subsequently, PRE represents a significant amount of baseline root exposure. The PRE as measured at 3 months represents 85% of the exposed root surface as measured at 12 months (Ghannad, 2007). Consequently, the additional root exposure represents a relatively small proportions of the overall root exposure and the additional root exposure that may be 75 attributed to the experimental treatment becomes diluted by the PRE, complicating the statistical analysis and identification of treatment associated differences. A further confounding factor related to our technique is that it does not account for any infra-bony attachment loss. In health, prior to any pathologic alveolar bone loss, the alveolus is thin bucco-lingually near the CEJ and widens apically in the mouse mandible (Page, 1982). Periodontal inflammation will precipitate alveolar bone loss within a circumscribed distance from the area of inflammation, which in humans has been determined to be about 2 mm (Waerhaug, 1978). In mice, the periodontal inflammation zone of bone loss is likely to encompass the entire aspect of the bone lingual at the molar teeth in the coronal regions of the alveolus. However, as the inflammation progresses apically with advancing disease, the potential for developing infrabony defects is amplified as the width of the alveolus increases, at some point most likely extending beyond the effective range of the inflammatory lesion (Tal, 1984). Consequently our technique for measuring alveolar bone loss may under-represent the true extent of alveolar bone loss by not accounting for infra-crestal bone loss. This supposition seems to be supported by published scanning electron micrographs and high-definition radiographs showing crater defects interproximally and moat defects on the facial and lingual surfaces (Ghannad et al., 2008). The lack of a significant difference noted between alveolar bone loss assessed from the control K1436 over-expressing group and the control W/T group, support the contention that these mice have normal levels of protection from periodontal disease. The findings 76 that the control f36-Rescue mouse strain expresses significantly greater alveolar bone loss than all other groups of control mice is in contrast to the findings for epithelial migration, epithelial furcation invasion and inflammation, and definitely inconsistent with our hypothesis. These findings may result from internal breeding and isolation of this subset of transgenic WIT mice, which are generally produced from a small initial breeding colony. It is also conceivable that these mice may be subject to greater levels of tooth eruption than the other strains of mice. However, the contribution of exaggerated tooth eruption seems inconsistent with our findings because tooth eruption in mice compensates for ocelusal tooth wear of the molar teeth (Page, 1982). As the tooth wear was not significantly different between the W/T and the 136 Rescue strains and all mice were of similar ages (data not shown), occiusal wear associated tooth eruption, if it contributed to root exposure, would have been equally associated with both strains of mice. It may also be said that the introduction of the K14 gene promoter may have contributed to root exposure with as of yet unknown cellular or gene expression effects that could have resulted in an exaggerated alveolar bone loss. These and potentially other factors could have all contributed in an additive manner to explain why the 136 Rescue strain of mice expressed exaggerated alveolar bone loss while expressing a significant rescue in terms of epithelial migration and gingival inflammation. Further studies will be needed to clarify this seemingly inconsistent observation. Periodontal inflammation generally progresses from gingival inflammation, though unpredictably, and is a histologic hallmark of active periodontal attachment loss and bone loss (Kinane and Attstrom, 2005). Most current theories surrounding the osteoclast 77 mediated loss of alveolar bone purport that pro-inflammatory substances including bacterially liberated LPS and host liberated pro-inflammatory cytokines TNFa and IL-i, work concurrently by different cellular mechanisms to provoke osteoclast differentiation and activation (Katagiri and Takahashi, 2002). As inflammation is central to the progression of periodontal bone loss, most current mechanical (Claffey et al., 2004), pharmacotherapeutic (Reddy et al., 2003) and possible future periodontal therapeutic strategies (Serhan, 2008) focus on resolving inflammation as a means of preventing permanent attachment loss. In this respect, understanding the endogenous factors that reduce inflammation locally can give insight towards future therapeutic modalities. Active TGFI31 is a well-established anti-inflammatory cytokine (Wright et al., 2003). The importance of the anti-inflammatory properties of TGFIH become evident in observations that knock-out of genes coding for TGFI31 activator proteins show various levels of inflammation and TGFI31 knock-out mice develop diffuse inflammation at many organ systems, succumbing to premature death soon after birth (Yang et aL, 2007). Periodontally, locally blocking TGF3l activation induces hallmarks of periodontal disease in a rat model (Ghannad et al., 2008). Conversely, gingival health has been associated with high levels of local TGF3i synthesis (Dutzan et al., 2009). Histologically, inflammation was graded at the interproximal regions immediately subjacent to the gingival epithelium at interproximal sites to assess if levels of inflammation could be attributed to the presence or absence of the integrin t36 sub-unit. As expected, the W/T strain showed the lowest level of gingival inflammation in all groups while the 136-I- strain showed significantly higher levels of gingival inflammation. 78 These fmdings are consistent with previously published results (Ghannad et al., 2008) and consistent with the anti-inflammatory properties of activated TGFI31. Inflammation assessed in the gingiva of the K 14136 strain was significantly lower than the j36-/- strain, not being significantly different than the WIT strain. Again this supports previous findings that the transgenic construct has a similar periodontal phenotype to the WIT strain although we accept that protection from periodontal inflammation may be solely contributed to by the murine 136 subunit, not by the h136 subunit. The results for gingival inflammation of the 136 Rescue strain, while revealing a small trend to reduced levels, are not significantly different compared to levels of gingival inflammation in the 136-I- strain, possibly suggesting that our data set was underpowered to identify differences between the 136-I- and B6 Rescue groups. Periodontal breakdown develops following a sustained hyper-inflammatory response by the periodontium primarily to bacterial plaque, however this breakdown is unpredictable at the subject level (Haffajee and Socransky, 1986; Kinane and Attstrom, 2005; Lindhe et al., 1983). In order to better study the patho-physiology of periodontal diseases, predictable models of periodontal diseases have been established that reliably result in periodontal attachment loss primarily via creating a sustained inflammatory response, most often in non-human primates and dogs (Page, 1982). Uncontrolled diabetes creates a hyper-inflammatory state providing a good model for investigating which factors are correlated with periodontal disease progression (Lalla et al., 1998a; Lalla Ct al., 2001). 79 Most non-human primate studies of periodontal disease are limited by study size and power due to the number of animals that can feasibly be included in any one study. Furthermore, the use of these larger animals with longer breeding cycles is inherently limited by the difficulties in creating genetic knock-out strains or identif’ing susceptible groups. The use of small rodents, specifically mice, overcomes both these obstacles by increasing number of animals in any one sample and the relative ease of developing genetically modified strains which greatly enhances the ability to observe the physiologic effects of induced genetic deficiencies and, in a reverse manner, imply actions of those missing genes. Lalla et.al. have reported on a murine model of periodontal disease that uses uncontrolled streptozotocine-induced diabetes in conjunction with topical application Porphyromonas gingivalis (Pg) (Lalla et al., 1998a). Pg has been implicated as a putative periodontal pathogen by its strong association with periodontal lesions (Haffajee et al., 2006) as well as its numerous virulence factors including the lipopolysaccharide (LPS) and fimbriae activation of toll-like receptors (TLRs) and released proteases and gingipains that breakdown host proteins (Madianos et al., 2005). In our model, the decision to not incorporate local inoculation with Pg was made to avoid the introduction of a third independent variable, further to experimental diabetes absence of the (36 integrin. A third independent variable would have unnecessarily complicated statistical analysis. Furthermore, in published reports, the use of experimental diabetes alone was sufficient to induce significant increases in alveolar bone loss over 3 months, reinforcing our decision to not use of Pg inoculation (Lalla et al., 1998a). However, two notable 80 differences in their methodology were the use of very young animals compared to our much older animals (4-5 months old at sacrifice, 14-15 months old at sacrifice, respectively) and the use of only male animals where we used equal numbers of both genders. In contrast to the work of Lalla et.al., (Lalla et al., 1 998a), recent publication seems to suggest that uncontrolled experimental diabetes alone may provide sufficient inflammatory stimulus to initiate an exaggeration of periodontal disease in the absence of co-inflammatory factors (Gomes et al., 2009). In this rat model of periodontitis, while uncontrolled diabetic rats showed a trend to greater bone loss, only diabetic rats that had ligature-induced inflammation at the molar teeth exhibited significantly greater bone loss. It must be noted that their experimental period was relatively short at 30 days and no mention of the animal ages were disclosed. Experimental diabetes induction with streptozotocin was predictable and maintainable over the experimental period in all groups with the experimental groups showing overt subjective signs of diabetes (polydipsia and polyuria) within 14-20 days subsequent to experimental diabetes induction. The subsequent accumulation of AGEs in the gingiva of diabetic mice has been previously established (Lalla et al., 1 998a). In apparent contrast to our expectations, experimental diabetes did significantly increase the levels of epithelial migration, bone loss and gingival inflammation when comparing experimental diabetes verses control animals. However, when bone loss was evaluated subsequent to gender specific analysis, small but significant differences were observed between the experimental and control animals. Male animals seemed to respond in manner consistent with our hypothesis. In the 136-I- group, experimental diabetes did significantly increase 81 the level of bone loss observed in the experimental (36-I- group over the control (36-/ group. However, experimental diabetes did not significantly affect bone loss within the other strains of male mice. This result suggests that the pro-inflammatory combination of experimental diabetes and integrin (36 deficiency imposes a significant enough burden to tip the scale towards periodontal disease initiation, and supports our contention that these two factors may act as co-factors in the progression of periodontal disease. The other groups by their natural or transgenic reintroduction of the integrin (36 subunit may have enough anti-inflammatory protection imparted by natural or transgenic (36, to prevent periodontal bone loss. As discussed previously, the exaggerated alveolar bone loss seen in the (36 rescue strain may be reflective of selective and repeated internal breeding, unknown promoter activity or tooth eruption. Slightly increased levels of bone loss seen in the male K14(36 group may reflect inefficiencies in the cellular association of the h(36 and murine av integrin subunits, or less than efficient activation of latent TGFf31 of the avh(36 transgenic integrin heterodimer. This suggestion is consistent with observations of a similar rescue strain used in studies of lung inflammation where the h(36 integrin subunit transgenic rescue mouse did not completely prevent pen-bronchial inflammation (Huang et al., 1998b). This explanation is also consistent with the observation of trends to increased inflammation and epithelial migration in the K14(36 strain, and may also explain why the (36 Rescue strain did not exhibit a phenotype completely consistent with the W/T phenotype. Alveolar bone loss results seen in the female subset imply dramatically different results. In a consistent fashion, the female diabetic groups of all strains exhibited less bone loss, 82 with the WIT, (36-I- and the (36 Rescue strains all exhibiting significant differences (P value <0.05, Student t-test). While these results are contrary to our expected results, some published research may help explain these observations. At about 15 months of age, the female mice were at risk of becoming anovulatory entering a period of reproductive life somewhat analogous to menopause in human females (Felicio et al., 1984; Gosden et al., 1983; Nelson et al., 1982). In human females, estrogen deficiency has been well associated with osteoporosis and mounting research is suggesting that the reduced estrogen levels characteristic of menopause may be associated with increases in tooth loss and periodontal attachment loss (Lerner, 2006; Meisel et al., 2008), possibly caused by the combination of a pro-inflammatory phenotype and reduced bone density. Additionally, estrogen has recently been positively associated with increasing OPG synthesis in human periodontal ligament cells in vitro, possibly contributing to a bone protective effect prior to menopause (Liang et al., 2008). However, mounting evidence from clinical cross-sectional studies have reported that diabetic postmenopausal females exhibit increased bone mineral density when compared to a similar groups of non- diabetic women (Hadjidakis et al., 2009; Hadzibegovic et al., 2008; Sosa et al., 2009), with more recent evidence suggesting that pre-menopausal diabetic females may also benefit from greater bone density (Gupta et al., 2009). As previous studies suggest positive correlations between maxillary and mandibular bone density and skeletal bone mass (Kribbs et al., 1989; Lindh et al., 2004) it may be expected that the reported increases in bone mineral density associated with diabetes, referred to above, likely will increased jaw bone density thereby conferring a protective effect against periodontal bone loss. While the mechanisms for these observations are yet to be elucidated, it is 83 interesting to speculate if female diabetic anovulatory mice may be protected against periodontal bone loss. A novel finding in our study stemmed from the observation that experimental diabetes adversely affected the 136-I- mouse strain (29% survival) during the 4-month experimental period, differing statistically from the experimental W/T group (100% survival). Additionally, the experimental (diabetic) f36 Rescue strain exhibited comparable levels of premature death to the experimental (diabetic) 136-I- strain (33% survival), but failed to reach statistical significance, possibly due to reduced numbers of mice in this group. Gender was not identified as a potential factor in this analysis. All animals that experienced premature death visually exhibited significant weight loss, lethargy and ultimately were emaciated at death. Changes occurred relatively quickly and once noted, affected animals were dead within 7-10 days. This constellation of features is suggestive of an acute diabetic complication associated with hyperglycemic fluid loss, known in humans as diabetic ketoacidosis (DKA). DKA is characterized by osmotic diuresis, hypovolemia, glycosuria and the associated severe electrolyte losses and subsequent serum electrolyte imbalances (Magee and Bhatt, 2001). Symptomatically, DKA will present with weight loss secondary to volume depletion, lethargy and hypotension, ultimately ending in coma and death. It is interesting to speculate on potential causes of premature mortality of these diabetic 136-I- mice related to the reduced TGF131 activation, as this is the primary known deficit in this strain of mice. A significant complication of chronic diabetes mellitus is end-stage renal disease characterized by glomeruloscierosis and interstitial fibrosis (Wada et al., 2007). The 84 association of elevated TGFI31 activation paralleling the development of diabetic nephropathy has been well established (Sharma and Ziyadeh, 1995), but only recently has up-regulation of the 136 integrin subunit at the distal tubule been reported in association with acute and chronic tubulointerstitial inflammation, while 136 intergrin is not normally expressed in healthy kidneys (Trevillian et al., 2004). Supporting and seemingly tying these previous findings together has been the recent observation in a murine model of progressive glomerulonephritis, the Alport mouse, that antibody-mediated blockage of av136 integrin significantly inhibited fibrotic and inflammatory changes in the kidney tissues concomitant with reductions in tissue levels of active TGF131 (Hahm et al., 2007). Accepting that we have no histology of our mouse kidney tissues to date, extrapolating from the previous studies, we may expect to see little if any kidney fibrosis in the 136-/ strain weather diabetic or not owing to their reduced potential for integrin av136 activation of TGF131. While speculating, it may be that the ability to induce kidney interstitial fibrosis is a protective mechanism that limits renal fluid loss, preserving hemodynamic and electrolyte balance and thereby reducing the potential for acute and potential lethal DKA. The physiologic cost of kidney fibrosis permits a longer-term viability of the organism as kidney function is associated with significant reserve, allowing significant loss of function prior to any symptoms becoming apparent (Raja and Coletti, 2006). While we cannot rule out the potential that the deaths may have been caused by end-stage renal disease which is a common complication of uncontrolled diabetes, the potential for this to be the etiologic factor responsible for the death of these animals is unlikely as end stage renal disease is associated with fluid retention, the opposite of what we observed clinically (Proctor et al., 2005). Furthermore, end-stage 85 renal disease is a slow progressive disease, not characterized by the rapid progression we observed. End-stage renal disease is associated with kidney fibrosis, a finding that would be highly unexpected in animals that may be protected from renal fibrosis (Hahm et al., 2007). This suggestion is supported by similar levels of premature death observed in our diabetic 6 Rescue mouse strain. Based on the measurable rescue from gingival inflammation and epithelial migration exhibited by the 136 Rescue strain, it may be surmised that active av136 integrin is expressed in the ginginval tissues, though no direct evidence of that is available. Extrapolating forward, knowing that ccvI36 integrin is expressed in kidney tissue and associated with kidney fibrosis (Hahm et al., 2007), we may expect that the 136 Rescue strain to be protected from premature death associated with diabetes due to the reintroduction of the h136 integrin. However, healthy kidney tissues are naturally deficient in cytokeratin 14 (Sicinnider et al., 2005) which would prevent active avf36 integrin from being expressed in the kidney tissues of 136 Rescue mice, thereby preventing these mice from developing TGF131 induced kidney fibrosis. If the kidney fibrosis is a protective mechanism, the 136 Rescue strain would not be protected. This would be consistent with our observations of elevated rates of premature death in the diabetic 136 Rescue strain, possibly by DKA. Further analysis and research is needed to clarifSr the role av136 plays in the protection against premature death of diabetic mice. 86 4.1 Limitations of our Study The exaggerated loss of mice from two of our diabetic groups is of importance to the interpretation of our results. As we are unaware of the specific cause of the premature mortality for these animals, the fact that they were not included for tissue analysis in the experimental 136-I- group and possibly the experimental j36 Rescue group, implies that these groups no-longer represent a random sampling of the population being tested. The premature death of these animals, if related to exaggerated inflammation, would imply that we have included only the subset of animals that were resistant to inflammation. If this were the case and we lost all the animals with average to hyper-inflammatory phenotypes, the retained sample would be skewed towards a hypo-inflammatory phenotype, potentially explaining the less than expected periodontal disease in the f36-I- group and possibly the 136 Rescue strains. In retrospect, or as a potential consideration for future studies, all animals of the original group should have been sacrificed when they presented with overt signs of impending death and kept for assessment of the experimental parameters as measured for the term animals. The data could be normalized in relation to the duration of the experimental diabetes using regression analysis or other means. An obvious limitation to the interpretation of our results is the number of subjects retained in each group. Initially, our power calculations suggested that 10 subjects per group should provide sufficient data to achieve significance, however with some suggestion that gender may be an issue, we were forced to further divide our groups to test for gender differences in some of the analysis. Not only did this reduce our number 87 of subjects per group, it also increased the number of inter-group comparisons. The number of comparisons made in any analysis of a single data set increases the potential error that is necessarily accommodated for in the post hoc adjustments with ANOVA testing (Altman, 1991). This serves to increase the differences required between groups to achieve significance. A further limitation for our data set was the number of sites evaluated histologically. While 47% of the tissue samples were sectioned and evaluated, we have a significant portion of un-evaluated data, which if included, would have increased our data size and possibly made a contribution to achieving statistical differences between groups. Some may argue that the methodology employed where the same subject provides multiple measurements and each measurement included as a discrete data point in the final comparison, is a violation of the experimental protocol and statistical limitations. However, periodontal disease has been repeatedly established as a site-specific disease, with individuals concurrently expressing areas of progression adjacent to areas of healing, supporting the use of site specific measurements as opposed to measurements at the level of the individual (Claffey and Egelberg, 1995; Haffajee and Socransky, 1986; Lindhe et al., 1983; Loesche and Grossman, 2001; Nunn and Harrel, 2001). The only credence to this contention, with respect to our research, is that much of the published research has been focused at looking for site-specific differences as etiologic factors while in our research the factors are based on genetic variability at the level of the subject. However, even high-risk individuals (probably genetically predisposed to periodontal disease) express concurrent areas of disease and resolution, further 88 supporting the use of the site as opposed to the individual as the level of measurement (Haffajee and Socransky, 1986). Considering the level of attention that has been devoted to discussing our results surrounding the 136 Rescue and K14136 strains, it is interesting to note that we have to date only little information on how the transgenes are affecting av136 integrin formation and activity on a cellular level. Use of 136 transgenic rescue strains in studies of lung inflammation has validated the proposed model of integrin av(36 activation of TGF(31 (Huang et al., 1998b). However, much of the purported activity of the transgenic integrin cxvh(36 is indirect evidence associated with levels of inflammation or quantifring active TGF(31 levels (Hakkinen et al., 2004). Little if any work has been published on the dimerization of the transgenic integrin heterodimer in the rescue strains and its level of activity compared to normal endogenous ctvf36 integrin in the strains specifically used in our study. The less than complete rescue of the f36 Rescue strain in terms of measures of inflammation, and epithelial migration suggest that the h(36 transgene expression is sub- optimal or spatio-temporally differentiated from the expression of the av gene consequently limiting the ability to form the av(36 heterodimer or, alternatively, the ability of the transgenic avh(36 heterodimer to activate TGF(31 is sub-optimal. While our hypothesis rests on the contention that integrin av(36 activates TGF13 which in turn acts to reduce inflammation locally, we have no direct evidence of levels of activity of TGF(31 or other cytokine levels within the gingival tissues for any of our animals. Histologic analysis of frozen sections of gingival tissues using immunohistochemistry techniques may give some insight to localizations of active TGF(31 and integrin cxvf36. Additionally, 89 tissue analysis using immunoblotting may give some information regarding cytokine and integrin levels possibly providing some information to the level of expression of the transgenic 136 integrin. Future studies may help clarify these questions and the interpretation of our results. 90 CHAPTER 5: CONCLUSIONS These results support our hypothesis and the recently published results that the loss of 136 integrin is associated with the initiation and progression of periodontal disease in a murine model through promoting a pro-inflammatory phenotype (Ghannad et al., 2008). The role of f36 integrin in periodontal health is reinforced by consistent trends in our findings that reductions in periodontal epithelial migration and inflammation are associated with reintroduction of the 136 integrin under a different promoter. We observed that diabetic male 136-I- mice exhibit significantly greater alveolar bone loss than non-diabetic 136-I- mice, while experimental diabetes in the other male strains failed to exaggerate alveolar bone loss. Furthermore, the modest reduction in epithelial migration with the 136 rescue phenotype implies that the periodontal disease associated with the 136-I- strain was not caused by the concurrent loss of other genes associated with the 136 gene knock-out. Evidence is mounting that integrin av136 is associated with activation of latent TGFf31, a cytokine associated with a hypo-inflammatory phenotype and protective to the periodontium if adequately activated. Conversely, individual variations in genotype associated with TGFI31 activation or activator proteins may have some association with genetic predisposition to periodontal diseases. The observation of significantly greater mortality associated with uncontrolled diabetes and the concurrent deficiency of the integrin 136 subunit, is to our knowledge, a novel finding and deserves further investigation to identifS’ the etiology and potential relevance to clinical practice. 91 5.1 Recommendations for Future Studies This study supports the protective nature of avf36 integrin in the junctional epithelium. However, a better understanding of the mechanisms involved with this integrin’s down- regulation during periodontal disease progression may offer some clues to the unpredictability that seems to be consistently associated with periodontal disease progression, even in the face of seemingly similar local inflammatory factors. Future studies may involve culturing junctional epithelium cells and trying to identify which factors are associated with reduced expression of av136 that accompanies a shift to pocket epithelium. Considering the novel observation associated with the mortality in the 136-f- diabetic •animals, further examinations should be undertaken to elucidate the actual pathophysiology of their rapid physical decline and premature death compared to the other strains. 92 References Acosta JB, del Barco DG, Vera DC, Savigne W, Lopez-Saura P, Guillen Nieto G, et al. (2008). The pro-inflammatory environment in recalcitrant diabetic foot wounds. mt WoundJ 5(4).53 0-9. Abmed N, Pansino F, Clyde R, Murthi P, Quinn MA, Rice GE, et al. (2002). Overexpression of alpha(v)beta6 integrin in serous epithelial ovarian cancer regulates extracellular matrix degradation via the plasminogen activation cascade. Carcinogenesis 23(2):237-44. Alberti KG, Zimmet PZ (1998). Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus provisional report of a WHO consultation. Diabet Med 15(7): 53 9-53. Altman DG (1991). Practical Statistics for Medical Research. London: Chapman & Hall, pp. 403-405. Aluwihare P, Munger JS (2008). What the lung has taught us about latent TGF-beta activation. Am JRespir Cell Mo! Biol 39(5):499-502. Annes JP, Rificin DB, Munger JS (2002). The integrin alphaVbeta6 binds and activates latent TGFbeta3. FEBSLett 51 1(i-3):65-8. Annes JP, Munger JS, Rifkin DB (2003). Making sense of latent TGFbeta activation. J Cell Sci ii 6(Pt 2):2 17-24. Annes JP, Chen Y, Munger JS, Rificin DB (2004). Integrin alphaVbeta6-mediated activation of latent TGF-beta requires the latent TGF-beta binding protein-i. J Cell Biol 165(5):723-34. Armitage GC (1999). Development of a classification system for periodontal diseases and conditions. Ann Periodontol 4(1): 1-6. Bampton IL, Shirlaw PJ, Topley S, Weller P, Wilton JM (1991). Human junctional epithelium: demonstration of a new marker, its growth in vitro and characterization by lectin reactivity and keratin expression. J Invest Dermatol 96(5):708- 17. Bartold PM, Walsh U, Narayanan AS (2000). Molecular and cell biology of the gingiva. Periodontol 2000 24:28-55. Baxter AG, Mandel TE (1991). Accelerated diabetes in non-obese diabetic (NOD) mice differing in incidence of spontaneous disease. Clin Exp Immunol 85(3).464-8. Berglundh T, Donati M (2005). Aspects of adaptive host response in periodontitis. J Clin Periodontol 32 Suppl 6:87-107. 93 Berrier AL, Yamada KM (2007). Cell-matrix adhesion. J Cell Physiol 21 3(3):565-73. Bonecchi R, Bianchi G, Bordignon PP, D’Ambrosio D, Lang R, Borsatti A, et al. (1998). Differential expression of chemokine receptors and chemotactic responsiveness of type 1 T helper cells (This) and Tb2s. JExp Med 187(1):129-34. Bosshardt DD, Lang NP (2005). The junctional epithelium: from health to disease. JDent Res 84(1):9-20. Breuss JM, Gallo J, DeLisser HM, Klimanskaya IV, Folkesson HG, Pittet if, et al. (1995). Expression of the beta 6 integrin subunit in development, neoplasia and tissue repair suggests a role in epithelial remodeling. J Cell Sd 108 (Pt 6):224 1-51. Brosius F (2003). Low-Dose Streptozotocine Induction Protocol (Mouse). Animal Models of Diabetic Complications Consortium Brown KA, Pietenpol JA, Moses ilL (2007). A tale of two proteins: differential roles and regulation of Smad2 and Smad3 in TGF-beta signaling. JCe11 Biochem 101(1):9-33. Calderwood DA (2004). Integrin activation. J Cell Sd 11 7(Pt 5): 657-66. Carranza J, Perez-Barberia FJ (2007). Sexual selection and senescence: male size dimorphic ungulates evolved relatively smaller molars than females. Am Nat 170(3):370- 80. Carro OM, Evans SA, Leone CW (1997). Effect of inflammation on the proliferation of human gingival epithelial cells in vitro. JPeriodontol 68(11):1070-5. Chapple IL (2009). Potential mechanisms underpinning the nutritional modulation of periodontal inflammation. JAm Dent Assoc 140(2): 178-84. Chavakis T, Bierhaus A, Al-Fakhri N, Schneider D, Wife 5, Linn T, et al. (2003). The pattern recognition receptor (RAGE) is a counterreceptor for leukocyte integrins: a novel pathway for inflammatory cell recruitment. JExp Med 198(10):1507-15. Chavakis T, Bierhaus A, Nawroth PP (2004). RAGE (receptor for advanced glycation end products): a central player in the inflammatory response. Microbes Infect 6(13):1219- 25. Chu PG, Weiss LM (2002). Keratin expression in human tissues and neoplasms. Histopathology 40(5):403-39. Claffey N, Egelberg J (1995). Clinical indicators of probing attachment loss following initial periodontal treatment in advanced periodontitis patients. J Clin Periodontol 22(9): 690-6. 94 Claffey N, Polyzois I, Ziaka P (2004). An overview of nonsurgical and surgical therapy. Periodontol 2000 36(35-44. Coulombe PA, Omary MB (2002). ‘Hard’ and ‘soft’ principles defining the structure, function and regulation of keratin intermediate filaments. Curr Opin Cell Biol 14(1): 110- 22. Crawford SE, Steilmach V, Murphy-Ulfrich JE, Ribeiro SM, Lawler J, Hynes RO, et al. (1998). Thrombospondin-1 is a major activator of TGF-betal in vivo. Cell 93(7). 1159- 70. Dallas SL, Sivakumar P, Jones CJ, Chen Q, Peters DM, Mosher DF, et al. (2005). Fibronectin regulates latent transforming growth factor-beta (TGF beta) by controlling matrix assembly of latent TGF beta-binding protein-i. JBiol Chem 280(19): 18871-80. De Martinis M, Di Benedetto MC, Mengoli LP, Ginaldi L (2006). Senile osteoporosis: is it an immune-mediated disease? Inflamm Res 55(10):399-404. Dutzan N, Gamonal J, Silva A, Sanz M, Vernal R (2009). Over-expression of forkhead box P3 and its association with receptor activator of nuclear factor-kappa B ligand, interleukin (IL) -17, IL- 10 and transforming growth factor-beta during the progression of chronic periodontitis. J Clin Periodontol 36(5):396-403. Engler AJ, Sen 5, Sweeney HL, Discher DE (2006). Matrix elasticity directs stem cell lineage specification. Cell 1 26(4):677-89. Eslami A, Gallant-Behm CL, Hart DA, Wiebe C, Honardoust D, Gardner H, et al. (2009). Expression of Integrin {alpha}v{beta}6 and TGF-{beta} in Scarless vs Scar-forming Wound Healing. JHistochem Cytochem (in press). Faria-Almeida R, Navarro A, Bascones A (2006). Clinical and metabolic changes after conventional treatment of type 2 diabetic patients with chronic periodontitis. J Periodontol 77(4):59i-8. Felicio LS, Nelson JF, Finch CE (1984). Longitudinal studies of estrous cyclicity in aging C57BL/6J mice: II. Cessation of cyclicity and the duration of persistent vaginal cornification. BiolReprod 31(3):446-53. Fiorellini JP KD, Ishikawa SO (2007). Chapter 4: The gingiva. In: Carranza’s Clinical Periodontology. M Newman, Takei, HH, Klokkevold PR, Carranza FA editor, pp. 48-57. Fujiwara N, Kobayashi K (2005). Macrophages in inflammation. Curr Drug Targets Inflamm Allergy 4(3):281-6. 95 Garguilo AW, Wentz F. M., Organ B. (1961). Dimensions of the Dentogingival Junction in Humans. JPeriodontol 32(July):26 1-267. Gemmell E, Yamazaki K, Seymour GJ (2002). Destructive periodontitis lesions are determined by the nature of the lymphocytic response. Crit Rev Oral Biol Med 13(1):17- 34. Ghannad F (2007). Expression of alphavbeta6 Integrin in Periodontal Disease. Vancouver, British Columbia, Canada, University of British Columbia. Ghannad F, Nica D, Fulle MI, Grenier D, Putnins EE, Johnston S, et al. (2008). Absence of alphavbeta6 integrin is linked to initiation and progression of periodontal disease. Am J Pathol 172(5): 1271-86. Goldin A, Beckman JA, Schmidt AM, Creager MA (2006). Advanced glycation end products: sparking the development of diabetic vascular injury. Circulation 114(6):597- 605. Gomes DA, Spolidorio DM, Pepato MT, Zuza EP, de Toledo BE, Goncalves A, et al. (2009). Effect of induced diabetes mellitus on alveolar bone loss after 30 days of ligature- induced periodontal disease. Jmt Acad Periodontol 11(2): 188-92. Gomis RR, Alarcon C, He W, Wang Q, Seoane J, Lash A, et al. (2006). A FoxO-Smad synexpression group in human keratinocytes. Proc NatlAcadSci USA 103(34):12747- 52. Gosden RG, Laing SC, Felicio LS, Nelson JF, Finch CE (1983). Imminent oocyte exhaustion and reduced follicular recruitment mark the transition to acyclicity in aging C57BL/6J mice. Biol Reprod 28(2).255-60. Graves DT, Fine D, Teng YT, Van Dyke TE, Hajishengallis G (2008). The use of rodent models to investigate host-bacteria interactions related to periodontal diseases. J C/in Periodontol 35(2):89- 105. Grossi SG, Skrepcinski FB, DeCaro T, Robertson DC, Ho AW, Dunford RG, et al. (1997). Treatment of periodontal disease in diabetics reduces glycated hemoglobin. J Periodontol 68(8).7 13-9. Grossi SG, Genco RJ (1998). Periodontal disease and diabetes mellitus: a two-way relationship. Ann Periodontol 3(1):51-61. Gupta R, Mohammed AM, Mojiminiyi OA, Alenizi EK, Abdulla NA (2009). Bone mineral density in premenopausal Arab women with type 2 diabetes mellitus. J Clin Densitom 12(1):54-7. 96 Gurkan A, Cinarcik S, Huseyinov A (2005). Adjunctive subantimicrobial dose doxycycline: effect on clinical parameters and gingival crevicular fluid transforming growth factor-beta levels in severe, generalized chronic periodontitis. J Clin Periodontol 32(3):244-53. Haapasalmi K, Zhang K, Tonnesen M, Olerud J, Sheppard D, Salo T, et al. (1996). Keratinocytes in human wounds express alpha v beta 6 integrin. Jlnvest Dermatol 1 06(1):42-8. Hadjidakis DI, Androulakis, II, Mylonakis AM, Sfakianakis ME, Raptis AE, Papaefstathiou AG, et al. (2009). Diabetes in postmenopause: different influence on bone mass according to age and disease duration. Exp Clin Endocrinol Diabetes 117(5): 199- 204. Hadzibegovic I, Miskic B, Cosic V, Prvulovic D, Bistrovic D (2008). Increased bone mineral density in postmenopausal women with type 2 diabetes mellitus. Ann Saudi Med 28(2): 102-4. Haffajee AD, Socransky SS (1986). Frequency distributions of periodontal attachment loss. Clinical and microbiological features. JClin Periodontol 13(6):625-37. Haffajee AD, Teles RP, Socransky SS (2006). The effect of periodontal therapy on the composition of the subgingival microbiota. Periodontol 2000 42:219-58. Hahm K, Lukashev ME, Luo Y, Yang WJ, Dolinski BM, Weinreb PH, et al. (2007). Alphav beta6 integrin regulates renal fibrosis and inflammation in Alport mouse. Am J Pathol 170(1).1 10-25. Hakkinen L, Westermarck J, Kahari VM, Larjava H (1996). Human granulation-tissue fibroblasts show enhanced proteoglycan gene expression and altered response to TGF beta 1. JDentRes 75(10):1767-78. Hakkinen L, Hildebrand HC, Berndt A, Kosmehl H, Larjava H (2000). Immunolocalization of tenascin-C, alpha9 integrin subunit, and alphavbeta6 integrin during wound healing in human oral mucosa. JHistochem Cytochem 48(7):985-98. Hakkinen L, Koivisto L, Gardner H, Saarialho-Kere U, Carroll JM, Lakso M, et al. (2004). Increased expression of beta6-integrin in skin leads to spontaneous development of chronic wounds. Am JPathol 164(1 ):229-42. Hakkinen L, Csiszar A (2007). Hereditary gingival fibromatosis: characteristics and novel putative pathogenic mechanisms. JDent Res 86(1):25-34. Hamidi S, Salo T, Kainulainen T, Epstein J, Lerner K, Larjava H (2000). Expression of alpha(v)beta6 integrin in oral leukoplakia. Br JCancer 82(8): 1433-40. 97 Harburger DS, Bouaouina M, Calderwood DA (2009). Kindlin-1 and -2 directly bind the C-terminal region of beta integrin cytoplasmic tails and exert integrin-specific activation effects. JBiol Chem. Harburger DS, Calderwood DA (2009). Integrin signalling at a glance. JCeJZ Sd 122(Pt 2): 159-63. Hatakeyama S, Yaegashi T, Oikawa Y, Fujiwara H, Mikami T, Takeda Y, et al. (2006). Expression pattern of adhesion molecules in junctional epithelium differs from that in other gingival epithelia. JPeriodontal Res 41(4):322-8. Horan GS, Wood S, Ona V, Li DJ, Lukashev ME, Weinreb PH, et al. (2008). Partial inhibition of integrin alpha(v)beta6 prevents pulmonary fibrosis without exacerbating inflaniniation. Am JRespir Grit Care Med 177(1):56-65. Hormia M, Owaribe K, Virtanen 1(2001). The dento-epithelial junction: cell adhesion by type I hemidesmosomes in the absence of a true basal lamina. JPeriodontol 72(6):788- 97. Huang X, Wu J, Spong S, Sheppard D (1998a). The integrin alphavbeta6 is critical for keratinocyte migration on both its known ligand, fibronectin, and on vitronectin. J Cell Sd 111 (Pt 15):2189-95. Huang X, Wu J, Zhu W, Pytela R, Sheppard D (1998b). Expression of the human integrin beta6 subunit in alveolar type II cells and bronchiolar epithelial cells reverses lung inflammation in beta6 knockout mice. Am JRespir Cell Mol Biol 19(4):636-42. Huang XZ, Wu SF, Cass D, Erle DJ, Cony D, Young SG, et al. (1996). Inactivation of the integrin beta 6 subunit gene reveals a role of epithelial integrins in regulating inflammation in the lung and skin. JCellBiol 133(4):921-8. Humphries ID, Byron A, Humphries MJ (2006). Integrin ligands at a glance. JCe11 Sci 119(Pt 19):3901-3. Hunter N, Nicholls B, Srivastava M, Chapple CC, Zoeliner HF, Gibbins JR (2001). Reactive pocket epithelium in untreated chronic periodontal disease: possible derivation from developmental remnants of the enamel organ and root sheath. J Oral Pathol Med 30(3): 178-86. Hynes RO (2002). Integrins: bidirectional, allosteric signaling machines. Cell 1 10(6):673-87. Ingber JS, Rose LF, Coslet JG (1977). The “biologic width”--a concept in periodontics and restorative dentistry. Alpha Omegan 70(3):62-5. 98 Itoh S, ten Dijke P (2007). Negative regulation of TGF-beta receptor/Smad signal transduction. Curr Opin Cell Biol 19(2):176-84. Janes SM, Watt FM (2004). Switch from alphavbeta5 to alphavbeta6 integrin expression protects squamous cell carcinomas from anoikis. J Cell Biol 1 66(3):4 19-31. Jenkins MK, Khoruts A, Ingulli B, Mueller DL, McSorley SJ, Reinhardt RL, et al. (2001). In vivo activation of antigen-specific CD4 T cells. Annu Rev Immunol 19:23-45. Jiang Q, Li DY (2005). [Cytokeratin expression in human junctional epithelium, oral epithelium and sulcular epithelium]. Zhonghua Kou Qiang Yi Xue Za Zhi 40(4):298-3 01. Jin U, Chiu GK, Corbet EF (2003). Are periodontal diseases risk factors for certain systemic disorders--what matters to medical practitioners? Hong Kong MedJ9(1):31-7. Jobling ME, Moff JD, Finnegan MT, Jurukovski V, Erickson AC, Walian PJ, et al. (2006). Isoform-specific activation of latent transforming growth factor beta (LTGF-beta) by reactive oxygen species. Radiat Res 166(6):839-48. Jones J, Watt FM, Speight PM (1997). Changes in the expression of alpha v integrins in oral squamous cell carcinomas. J Oral Pathol Med 26(2):63-8. Jones IL, Walker RA (1999). Integrins: a role as cell signalling molecules. Mo! Pathol 52(4):208- 13. Kantarci A, Van Dyke TE (2005). Resolution of inflammation in periodontitis. J Periodonto! 76(11 Suppl):2 168-74. Karring T, Lang NP, Loe H (1975). The role of gingival connective tissue in determining epithelial differentiation. J Periodontal Res 10(1): 1-11. Katagiri T, Takahashi N (2002). Regulatory mechanisms of osteoblast and osteoclast differentiation. Oral Dis 8(3): 147-59. Katz J, Bhattacharyya I, Farkhondeh-Kish F, Perez FM, Caudle Rivi, Heft MW (2005). Expression of the receptor of advanced glycation end products in gingival tissues of type 2 diabetes patients with chronic periodontal disease: a study utilizing immunohistochemistry and RT-PCR. J Clin Periodontol 32(1):40-4. Katz J, Yoon TY, Mao 5, Lamont RJ, Caudle RM (2007). Expression of the receptor of advanced glycation end products in the gingival tissue of smokers with generalized periodontal disease and after nornicotine induction in primary gingival epithelial cells. J Periodontol 78(4). 736-41. Kavanagh KD, Evans AR, Jernvall J (2007). Predicting evolutionary patterns of mammalian teeth from development. Nature 449(7161):427-32. 99 Keski-Oja J, Koli K, von Melchner H (2004). TGF-beta activation by traction? Trends Cell Biol 14(12):657-9. Kinane DF, Attstrom R (2005). Advances in the pathogenesis of periodontitis. Group B consensus report of the fifth European Workshop in Periodontology. J Clin Periodontol 32 Suppi 6:130-1. Kiran M, Arpak N, Unsal E, Erdogan MF (2005). The effect of improved periodontal health on metabolic control in type 2 diabetes mellitus. J Clin Periodontol 32(3):266-72. Kobayashi K, Rose GG, Mahan CJ (1976). Ultrastructure of the dento-epithelial junction. JPeriodontal Res 1 1(6):3 13-30. Kribbs PJ, Chesnut CH, 3rd, Ott SM, Kilcoyne RF (1989). Relationships between mandibular and skeletal bone in an osteoporotic population. JProsthet Dent 62(6):703-7. Kuilcarni AB, Huh CG, Becker D, Geiser A, Lyght M, Flanders KC, et al. (1993). Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc NatlAcadSci USA 90(2):770-4. Lalla E, Lamster TB, Feit M, Huang L, Schmidt AM (1998a). A murine model of accelerated periodontal disease in diabetes. J Periodontal Res 33(7): 387-99. Lalla E, Lamster TB, Schmidt AM (1998b). Enhanced interaction of advanced glycation end products with their cellular receptor RAGE: implications for the pathogenesis of accelerated periodontal disease in diabetes. Ann Periodontol 3(1): 13-9. Lalla E, Lamster IB, Feit M, Huang L, Spessot A, Qu W, et al. (2000). Blockade of RAGE suppresses periodontitis-associated bone loss in diabetic mice. J Clin Invest 105(8):1 117-24. Lalla E, Lamster IB, Stern DM, Schmidt AM (2001). Receptor for advanced glycation end products, inflammation, and accelerated periodontal disease in diabetes: mechanisms and insights into therapeutic modalities. Ann Periodontol 6(1): 113-8. Lalla E, Cheng B, La! S, Kaplan 5, Softness B, Greenberg E, et a!. (2007). Diabetes mellitus promotes periodontal destruction in children. J Clin Periodontol 34(4):294-8. Lamers ML, Gimenes FA, Nogueira FN, Nicolau J, Gama P, Santos MF (2007). Chronic hyperglycaemia increases TGFbeta2 signaling and the expression of extracellular matrix proteins in the rat parotid gland. Matrix Biol 26(7):572-82. Larjava H, Plow EF, Wu C (2008). Kindlins: essential regulators of integrin signalling and cell-matrix adhesion. EMBO Rep 9(12): 1203-8. 100 Legate KR, Wickstrom SA, Fassler R (2009). Genetic and cell biological analysis of integrin outside-in signaling. Genes Dev 23(4):397-41 8. Lekic PC, Pender N, McCulloch CA (1997). Is fibroblast heterogeneity relevant to the health, diseases, and treatments of periodontal tissues? Cr1! Rev Oral Biol Med 8(3):253- 68. Lerner UH (2006). Inflammation-induced bone remodeling in periodontal disease and the influence of post-menopausal osteoporosis. J Dent Res 85(7): 596-607. Li MO, Wan YY, Sanjabi S, Robertson AK, Flavell RA (2006). Transforming growth factor-beta regulation of immune responses. Annu Rev Immunol 24:99-146. Liang L, Yu IF, Wang Y, Ding Y (2008). Estrogen regulates expression of osteoprotegerin and RANKL in human periodontal ligament cells through estrogen receptor beta. JPeriodontol 79(9): 1745-51. Like AA, Rossini AA (1976). Streptozotocin-induced pancreatic insulitis: new model of diabetes mellitus. Science 193(4251 ):4 15-7. Lindh C, Obrant K, Petersson A (2004). Maxillary bone mineral density and its relationship to the bone mineral density of the lumbar spine and hip. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 98(1): 102-9. Lindhe J, Haffajee AD, Socransky SS (1983). Progression of periodontal disease in adult subjects in the absence of periodontal therapy. J Clin Periodontol 1 0(4):433-42. Locke M, Hyland PL, Irwin CR, Mackenzie IC (2008). Modulation of gingival epithelial phenotypes by interactions with regionally defined populations of fibroblasts. J Periodontal Res 43(3):279-89. Loe H, Theilade E, Jensen SB (1965). Experimental Gingivitis in Man. JPeriodontol 36:177-87. Loe H, Anerud A, Boysen H, Morrison E (1986). Natural history of periodontal disease in man. Rapid, moderate and no loss of attachment in Sri Lankan laborers 14 to 46 years of age. J Clin Periodontol 1 3(5):43 1-45. Loe H (1993). Periodontal disease. The sixth complication of diabetes mellitus. Diabetes Care 16(1):329-34. Loesche WJ, Grossman NS (2001). Periodontal disease as a specific, albeit chronic, infection: diagnosis and treatment. Clin Microbiol Rev 14(4):727-52. Loos BG (2005). Systemic markers of inflammation in periodontitis. JPeriodontol 76(11 Suppl):2106-15. 101 Luo BH, Carman CV, Springer TA (2007). Structural basis of integrin regulation and signaling. Annu Rev Immunol 25:619-47. Ma U, Yang H, Gaspert A, Carlesso G, Barty MM, Davidson JM, et al. (2003). Transforming growth factor-beta-dependent and -independent pathways of induction of tubulointerstitial fibrosis in beta6(-/-) mice. Am JPathol 163(4): 1261-73. Madianos PN, Bobetsis YA, Kinane DF (2005). Generation of inflammatory stimuli: how bacteria set up inflammatory responses in the gingiva. J Clin Periodontol 32 Suppi 6:57- 71. Magee MF, Bhatt BA (2001). Management of decompensated diabetes. Diabetic ketoacidosis and hyperglycemic hyperosmolar syndrome. Crit Care Clin 17(1):75-106. Mealey BL, Oates TW (2006). Diabetes mellitus and periodontal diseases. JPeriodontol 77(8): 1289-303. Mealey BL, Ocampo GL (2007). Diabetes mellitus and periodontal disease. Periodontol 2000 44:127-53. Meisel P, Reifenberger J, Haase R, Nauck M, Bandt C, Kocher T (2008). Women are periodontally healthier than men, but why don’t they have more teeth than men? Menopause 1 5(2):270-5. Muller-Glauser W, Schroeder HE (1982). The pocket epithelium: a light- and electronmicroscopic study. JPeriodontol 53(3): 133-44. Munger JS, Huang X, Kawakatsu H, Griffiths MJ, Dalton SL, Wu J, et a!. (1999). The integrin alpha v beta 6 binds and activates latent TGF beta 1: a mechanism for regulating pulmonary inflammation and fibrosis. Cell 96(3):319-28. Nanci A, Bosshardt DD (2006). Structure of periodontal tissues in health and disease. Periodontol 2000 40:11-28. Nelson JF, Felicio LS, Randall PK, Sims C, Finch CE (1982). A longitudinal study of estrous cyclicity in aging C57BL/6J mice: I. Cycle frequency, length and vaginal cytology. Biol Reprod 27(2): 327-39. Nelson RG (2008). Periodontal disease and diabetes. Oral Dis 14(3):204-5. Niu Y, Xie T, Ge K, Lin Y, Lu 5 (2008). Effects of extracellular matrix glycosylation on proliferation and apoptosis of human dermal fibroblasts via the receptor for advanced glycosylated end products. Am J Dermatopathol 30(4):344-5 1. 102 Nunn ME, Harrel SK (2001). The effect of occiusal discrepancies on periodontitis. I. Relationship of initial occlusal discrepancies to initial clinical parameters. JPeriodontol 72(4).485-94. Offenbacher S, Katz V, Fertik G, Collins J, Boyd D, Maynor G, et al. (1996). Periodontal infection as a possible risk factor for preterm low birth weight. JPeriodontol 67:10 Suppl).1 103-13. Oksonen J, Sorokin LM, Virtanen, Hormia M (2001). The junctional epithelium around murine teeth differs from gingival epithelium in its basement membrane composition. J Dent Res 80(12):2093-7. Omary MB, Coulombe PA, McLean WH (2004). Intermediate filament proteins and their associated diseases. N Engi JMed 351 (20):2087- 100. Ouhayoun JP, SawafMH, Gofflaux JC, Etienne D, Forest N (1988). Re-epithelialization of a palatal connective tissue graft transplanted in a non-keratinized alveolar mucosa: a histological and biochemical study in humans. JPeriodontal Res 23(2): 127-33. Overall CM, Wrana JL, Sodek J (1991). Induction of formative and resorptive cellular phenotypes in human gingival fibroblasts by TGF-beta 1 and concanavalin A: regulation of matrix metalloproteinases and TIMP. JPeriodontal Res 26(3 Pt 2):279-82. Overman DO, Salonen 11(1994). Characterization of the human junctional epithelial cells directly attached to the tooth (DAT cells) in periodontal disease. JDent Res 73(12): 1818- 23. Page R, Schroeder, HE (1982). Periodontitis in man and other animals: a comparitive review New York: Basel. Page RC, Schroeder HE (1976). Pathogenesis of inflammatory periodontal disease. A summary of current work. Lab Invest 34(3):235-49. Page RC, Offenbacher S, Schroeder HE, Seymour GJ, Kornman KS (1997). Advances in the pathogenesis of periodontitis: summary of developments, clinical implications and future directions. Periodontol 2000 14:216-48. Piccinni MP (2006). T cells in normal pregnancy and recurrent pregnancy loss. Reprod Biomed Online 13(6):840-4. Pollanen MT, Salonen 11, Uitto VJ (2003). Structure and function of the tooth-epithelial interface in health and disease. Periodontol 2000 31:12-31. Prieto AL, Edelman GM, Crossin KL (1993). Multiple integrins mediate cell attachment to cytotactinltenascin. Proc Natl Acad Sci US A 90(21): 10154-8. 103 Prime SS, Pring M, Davies M, Paterson IC (2004). TGF-beta signal transduction in oro facial health and non-malignant disease (part I). Crit Rev Oral Biol Med 15(6):324-36. Pritlove-Carson S, Charlesworth S, Morgan PR, Palmer RM (1997). Cytokeratin phenotypes at the dento-gingival junction in relative health and inflammation, in smokers and nonsmokers. Oral Dis 3(1):19-24. Proctor R, Kumar N, Stein A, Moles D, Porter S (2005). Oral and dental aspects of chronic renal failure. JDent Res 84(3): 199-208. Rahimi RA, Leof EB (2007). TGF-beta signaling: a tale of two responses. J Cell Biochem 102(3):593-608. Raja K, Coletti DP (2006). Management of the dental patient with renal disease. Dent Clin North Am 50(4):529-45. Reddy MS, Geurs NC, Gunsolley JC (2003). Periodontal host modulation with antiproteinase, anti-inflammatory, and bone-sparing agents. A systematic review. Ann Periodontol 8(1): 12-37. Robertson AK, Hansson GK (2006). T cells in atherogenesis: for better or for worse? Arterioscier Thromb Vase Biol 26(1 1):2421-32. Saito T, Shimazaki Y, Kiyohara Y, Kato I, Kubo M, lida M, et al. (2004). The severity of periodontal disease is associated with the development of glucose intolerance in non- diabetics: the Hisayama study. JDent Res 83(6):485-90. Sampath TK, Muthukumaran N, Reddi AH (1987). Isolation of osteogenin, an extracellular matrix-associated, bone-inductive protein, by heparin affinity chromatography. Proc Natl Acad Sd USA 84(20):7 109-13. Schmidt R, Wirtala, J (1996). Modification of Movat Pentachrome Stain with Improved reliability of Elastin Staining. Journal ofHistochemistry 19(4): 325-327. Schrementi ME, Ferreira AM, Zender C, DiPietro LA (2008). Site-specific production of TGF-beta in oral mucosal and cutaneous wounds. Wound Repair Regen 16(1):80-6. Schroeder HE, Listgarten MA (1997). The gingival tissues: the architecture of periodontal protection. Periodontol 2000 13:91-120. Schroeder HE, Listgarten MA (2003). The junctional epithelium: from strength to defense. JDentRes 82(3):158-61. Seppala B, Ainamo J (1994). A site-by-site follow-up study on the effect of controlled versus poorly controlled insulin-dependent diabetes mellitus. J Clin Periodontol 21(3): 161-5. 104 Serhan CN (2008). Controlling the resolution of acute inflammation: a new genus of dual anti-inflammatory and proresolving mediators. JPeriodontol 79(8 Suppl): 1520-6. Seymour GJ, Gemmell E (2001). Cytokines in periodontal disease: where to from here? Acta Odontol Scand 59(3):167-73. Sharma K, Ziyadeh FN (1995). Hyperglycemia and diabetic kidney disease. The case for transforming growth factor-beta as a key mediator. Diabetes 44(10):1139-46. Sheppard D (2000). In vivo functions of integrins: lessons from null mutations in mice. Matrix Biol 19(3):203-9. Sheppard D (2001). Integrin-mediated activation of transforming growth factor-beta(1) in pulmonary fibrosis. Chest 120(1 Suppl):49S-53S. Sheppard D (2005). Integrin-mediated activation of latent transforming growth factor beta. Cancer Metastasis Rev 24(3): 395-402. Sheppard D (2006). Transforming growth factor beta: a central modulator of pulmonary and airway inflammation and fibrosis. Proc Am Thorac Soc 3(5):413-7. Shimono M, Ishikawa T, Enokiya Y, Muramatsu T, Matsuzaka K, Inoue T, et al. (2003). Biological characteristics of the junctional epithelium. JElectron Microsc (Tokyo) 52(6): 627-3 9. Skinnider BF, Folpe AL, Hennigar RA, Lim SD, Cohen C, Tamboli P, et al. (2005). Distribution of cytokeratins and vimentin in adult renal neoplasms and normal renal tissue: potential utility of a cytokeratin antibody panel in the differential diagnosis of renal tumors. Am JSurg Pathol 29(6):747-54. Sosa M, Saavedra P, Jodar E, Lozano-Tonkin C, Quesada JM, Torrijos A, et al. (2009). Bone mineral density and risk of fractures in aging, obese post-menopausal women with type 2 diabetes. The GIUMO Study. Aging Clin Exp Res 21(1):27-32. Szabo SJ, Sullivan BM, Peng SE, Glimcher LH (2003). Molecular mechanisms regulating Thi immune responses. Annu Rev Immunol 21:713-58. Tal H (1984). Relationship between the interproximal distance of roots and the prevalence of intrabony pockets. JPeriodontol 55(10):604-7. Taubman MA, Valverde P, Han X, Kawai T (2005). Immune response: the key to bone resorption in periodontal disease. JPeriodontol 76(11 Suppl):2033-41. Taylor GW, Borgnakke WS (2008). Periodontal disease: associations with diabetes, glycemic control and complications. Oral Dis 14(3):191-203. 105 Teng YT (2006). Protective and destructive immunity in the periodontium: Part 1--innate and humoral immunity and the periodontium. JDent Res 85(3):198-208. Theilade E, Wright WH, Jensen SB, Loe H (1966). Experimental gingivitis in man. II. A longitudinal clinical and bacteriological investigation. JPeriodontal Res 1:1-13. Thomas GJ, Nystrom ML, Marshall IF (2006). Alphavbeta6 integrin in wound healing and cancer of the oral cavity. J Oral Pathol Med 35(1): 1-10. Tonetti MS, Imboden MA, Gerber L, Lang NP, Laissue 3, Mueller C (1994). Localized expression of mRNA for phagocyte-specific chemotactic cytokines in human periodontal infections. Infect Immun 62(9):4005-14. Trevillian P, Paul H, Millar E, Hibberd A, Agrez MV (2004). alpha(v)beta(6) Integrin expression in diseased and transplanted kidneys. Kidney mt 66(4):1423-33. Tu YK, Gilthorpe MS, Griffiths GS, Maddick IH, Eaton KA, Johnson NW (2004). The application of multilevel modeling in the analysis of longitudinal periodontal data--part I: absolute levels of disease. JPeriodontol 75(1): 127-36. Uitto VJ, Larjava H (1991). Extracellular matrix molecules and their receptors: an overview with special emphasis on periodontal tissues. Crit Rev Oral Biol Med 2(3):323- 54. Vitkov L, Krautgartner WD, Hannig M (2005). Surface morphology of pocket epithelium. Ultrastruct Pathol 29(2): 121-7. Wada T, Sakai N, Matsushima K, Kaneko 5 (2007). Fibrocytes: a new insight into kidney fibrosis. Kidney mt 72(3):269-73. Waerhaug J (1978). Healing of the dento-epithelial junction following subgingival plaque control. I. As observed in human biopsy material. JPeriodontol 49(1):1-8. Wang J, Zhang Z, Xu K, Sun X, Yang G, Niu W, et al. (2008). Suppression of integrin alphaupsilonbeta6 by RNA interference in colon cancer cells inhibits extracellular matrix degradation through the MAPK pathway. IntJCancer 123(6): 1311-7. Watt FM (2002). Role of integrins in regulating epidermal adhesion, growth and differentiation. Embo J 21(15): 3919-26. Westfelt E, Rylander H, Blobme G, Jonasson P, Lindhe 3 (1996). The effect of periodontal therapy in diabetics. Results after 5 years. JClin Periodontol 23(2):92-100. 106 Wiebe CB, Adkins CA, Putnins EE, Hakkinen L, Larjava HS (2001). Naturally occurring periodontal bone loss in the wild deer mouse, genus Peromyscus. JPeriodontol 72(5):620-5. Willberg J, Syrjanen S, Honnia M (2006). Junctional epithelium in rats is characterized by slow cell proliferation. JPeriodontol 77(5): 840-6. Wipff PJ, Rifldn DB, Meister JJ, Hinz B (2007). Myofibroblast contraction activates latent TGF-betal from the extracellular matrix. JCe11 Biol 179(6): 1311-23. Wright HJ, Chapple IL, Matthews JR (2003). Levels of TGFbetal in gingival crevicular fluid during a 21-day experimental model of gingivitis. Oral Dis 9(2):88-94. Yang Z, Mu Z, Dabovic B, Jurukovski V, Yu D, Sung J, et al. (2007). Absence of integrin-mediated TGFbetal activation in vivo recapitulates the phenotype of TGFbetal null mice. J Cell Biol 176(6): 787-93. Young GD, Murphy-Ullrich JE (2004). Molecular interactions that confer latency to transforming growth factor-beta. JBiol Chem 279(36):38032-9. Zaidel-Bar R, Itzkovitz S, Ma’ayan A, Iyengar R, Geiger B (2007). Functional atlas of the integrin adhesome. Nat Cell Biol 9(8): 85 8-67. Zambruno G, Marchisio PC, Marconi A, Vaschieri C, Meichiori A, Giarmetti A, et al. (1995). Transforming growth factor-beta 1 modulates beta 1 and beta 5 integrin receptors and induces the de novo expression of the alpha v beta 6 heterodimer in normal human keratinocytes: implications for wound healing. J Cell Biol 129(3):853-65. Zhou Y, Koli K, Hagood JS, Miao M, Mavalli M, Rifkin DB, et a!. (2009). Latent transforming growth factor-beta-binding protein-4 regulates transforming growth factor- beta! bioavailability for activation by fibrogenic lung fibroblasts in response to bleomycin. Am JPathol 174(1 ):2 1-33. 107 APPENDIX https:f/rise.ubc.ca/ri/Oocf0/VO3JQPCS5I6KLAM?l... THE UNIVCRSErV OF RRTTTSH COLUMBIA FuedIn Agtncv: Funding Title: ANIMAL CARE CERTIFICATE Office of Research scrvicee and Administration 102, 6190 Agronomy Road, Vancouver, BC VET 123 Phone: 504-827 5111 Fax: 604422-509) Application Number: A07-0t95 ‘mvcstigatnr or Course Dl rector: l-taniu SLarjava Department: Oral hloIoç$cal & Medical Scisnces An iinels: Me J’VB 125 • Mlce betaó-f- t25 Start Date: December 1., 1007 March 5. 2009 Fund inq Sources: Canadian Institutes of Health Researtf, (CIHR) Fundhig Tltle Function of keratiriocyte lvBt Integrin in TCFB activation Canadian Institutes of Health Research (CIHR) Funcbon of 4ceratinocyte alphn-vbeo-6 integrin hi wound healing Unfunded tItle: N/A The Animal Care Committee has axemined and approved the use of nni male for the above expari mental project. This certiftate Is valid for one ‘year from the above start or approva date (whiciever Is later) provided there is no change in the exp&mental proceduros. Annual review is requireti by the CCAC and some granting agencies. A copy of this certificate roust be dIsplayed In your animal facility. 1 of 1 10/03/2009 10:25 PH 108


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items