Decellularized Tissue Engineered Constructs Using Cell Sheet Technology

Author Farag, Amro Ahmed Mahmoud

Published 2016

Thesis Type Thesis (PhD Doctorate)

School School of Dentistry and Oral Health

DOI https://doi.org/10.25904/1912/2310

Downloaded from http://hdl.handle.net/10072/367720

Griffith Research Online https://research-repository.griffith.edu.au Decellularized Tissue engineered constructs using cell sheet technology

Amro Ahmed Mahmoud Farag BDSc, M.Sc. (Periodontics) School of Dentistry Griffith Health

Griffith University Submitted in fulfilment of the requirements of the degree of Doctor of Philosophy

3/30/2016

I Statement of Originality

I, Amro Farag declare that this work has not previously been submitted for a degree or diploma in any university. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made in the thesis itself.

(Signed)______

Amro Farag

II Contents

Statement of Originality ...... II

Contents ...... III

List of Tables ...... VIII

List of Figures...... IX

Abstract ...... XI

Acknowledgements ...... XIII

Publications and successful grants arising from this thesis ...... XIV

Chapter 1 ...... 1

Introduction and Literature Review ...... 1

Introduction: ...... 1

Literature Review ...... 3

Structure of the periodontium: ...... 3

Effect of periodontal disease on periodontal tissues: ...... 6

Periodontal regeneration: ...... 7

Periodontal regenerative therapies: ...... 8

Guided tissue regeneration (GTR): ...... 9

Bone Grafting: ...... 12

Bioactive materials & signaling molecules: ...... 15

Tissue engineering of periodontal ligament: ...... 19

Cell sheet technology: ...... 21

Tissue engineered decellularized matrices and periodontal regeneration: ...... 22

Research hypothesis: ...... 26

Aims: ...... 26

Potential Significance and Innovation...... 27

Chapter 2 ...... 28

III Decellularized periodontal ligament cell sheets with recellularization potential ...... 28

Abstract ...... 29

Introduction ...... 30

Materials and methods ...... 31

Primary human periodontal ligament cells isolation and culture ...... 31

Decellularization, immunostaining and confocal imagining of cultured cell monolayers...... 31

Melt Electrospun carrier membrane fabrication ...... 32

Cell sheet harvesting ...... 33

Decellularization of cell sheets ...... 33

Confocal imaging of cell sheets ...... 34

DNA content ...... 34

Scanning electron microscopy (SEM) of cell sheets ...... 34

Growth factor ELISA assay ...... 35

Collagen quantification: ...... 35

Recellularization of decellularized sheets ...... 36

Statistical analysis...... 36

Results ...... 37

Scanning electron microscopy ...... 37

Extracellular matrix characterization ...... 38

DNA quantification ...... 39

Growth factors ELISA and collagen quantification ...... 39

Cell growth on decellularized sheets ...... 40

Discussion ...... 43

Chapter 3 ...... 47

Assessment of static and perfusion methods for decellularization of PCL membrane supported fibrous cell sheet constructs ...... 47

Abstract ...... 48

Introduction ...... 49

Materials and methods ...... 51

Membrane fabrication via melt electrospinning writing ...... 51

Primary human periodontal ligament cells (h-pdl cells) isolation and proliferation ...... 51

Cell sheet harvesting ...... 51

Decellularization protocols ...... 52

Confocal imaging of cell sheets ...... 56

Scanning electron microscopy of cell sheets ...... 56

Quantification of DNA ...... 57

Collagen quantification: ...... 57

Growth factor Bioplex assay ...... 58

Statistical analysis...... 59

Results ...... 59

DNA removal and DNase effect on decellularization ...... 59

Structural preservation of decellularized sheets ...... 61

Collagen content preservation ...... 64

Growth factors retention ...... 65

Discussion ...... 65

Chapter 4 ...... 69

The effect of decellularized periodontal ligament constructs on cell differentiation in vitro and periodontal regeneration in vivo...... 69

Abstract ...... 70

Introduction ...... 71

Materials and Methods ...... 72

Scaffold fabrication by melt electrospinning writing ...... 72

Covering barrier membrane ...... 73 V

Primary human periodontal ligament cells (h-pdl cells) isolation and proliferation ...... 73

Cell sheet harvesting for in vitro experiment ...... 73

Perfusion Decellularization ...... 74

Recellularization of decellularized sheets for gene expression and STRO-1 immunostaining ...... 74

RNA extraction and cDNA synthesis ...... 75

Gene expression analysis ...... 75

Human stromal precursor antigen-1 (STRO-1) immunostaining and confocal imaging ...... 76

In vivo study...... 76

Cell sheet preparation and decellularization ...... 76

Rat periodontal defect ...... 78

Microcomputed tomography (µCT) ...... 80

Hematoxylin and eosin (H&E) and Mason Trichome histology (MT) staining... 81

Immunohistochemistry ...... 81

Statistical analysis...... 82

Results ...... 82

Decellularised cell sheets induce bone markers gene expression in human placental mesenchymal stem cells and periodontal ligament cells ...... 82

Differentiation of placental mesenchymal stem cells initiated by the decellularized cell sheets shown by loss of STRO-1 expression...... 88

In vivo evaluation of the decellularized cell sheets in a rat mandibular defect .... 89

Discussion ...... 99

Chapter 5 ...... 103

Discussion and future directions ...... 103

Conclusion ...... 108

Future directions:...... 108

Perspective studies ...... 108

References ...... 109

Appendix ...... 140

Instruction manual for decellularized constructs fabrication and validation ...... 140

1. Materials ...... 140

1.1. Primary Human periodontal ligament cell (hPDLC) harvesting and expansion ...... 140

1.2. Melt electrospun PCL carrier membrane fabrication ...... 141

1.3. Cell sheet fabrication and harvesting...... 141

1.4. Perfusion decellularization components ...... 141

1.5. Cell sheet fixation and preparation for immunostaining for confocal imaging...... 141

1.6. Growth factor extraction ...... 142

2. Methods...... 142

2.1. Primary Human periodontal ligament cell (hPDLC) harvesting and expansion ...... 142

2.2. Melt electrospun PCL carrier membrane fabrication ...... 143

2.3. Cell sheet fabrication and harvesting...... 144

2.4. Cell sheet construct decellularization ...... 145

2.5. Immunostaining of cell sheet constructs ...... 146

2.6. Growth factor extraction ...... 147

3. Notes ...... 147

VII

List of Tables

Table 3.1. Decellularization methods used in this study ...... 53 Table 4.1: List of primers for gene expression study...... 77 Table 4.2: Summary of gene expression results...... 87

VIII List of Figures

Figure 2.1: Human periodontal ligament (HPDL) tissues explanted, HPDL cells expanded, cell sheet light microscopy before and after decellularization, melt electrospinning machine and polycaprolactone (PCL) scaffold fabrication and Perfusion device...... 36 Figure 2.2: Harvesting of fresh Human periodontal ligament (HPDL) cell sheet and scanning electron microscopy (SEM) showing fresh and decellularized PDL sheet. . 38 Figure 2.3: Immunostaining of human collagen type I and fibronectin...... 39 Figure 2.4: Comparison of DNA amounts, growth factor concentrations and collagen content of fresh and decellularized periodontal ligament cell sheet constructs...... 40 Figure 2.5: Recellularization potential of the decellularized sheet after seeding with allogenic hPDL cells after 3, 7 and 21 days respectively...... 41 Figure 2.6: Secondary antibody control for immunostaining. (A&B) Nuclei (DAPI) in blue, Actin filaments (Phalloidin) in red...... 42 Figure 2.7: Seeding of hPDL cells on BioGide collagen membrane over 3, 7 and 21 days respectively...... 42 Figure 2.8: SEM of multilayered (4 layers) hPDL sheets (A-C) hPDL multilayered cell sheets before and after decellularization...... 43 Figure 3.1. Decellularization perfusion system ...... 54 Figure 3.2. DNA, collagen and retained growth factor quantification in decellularized human periodontal ligament cell sheets...... 60 Figure 3.3. Confocal microscopy and SEM of fresh cell sheet...... 62 Figure 3.4. Confocal imaging and SEM of static decellularized cell sheets...... 63 Figure 3.5: Confocal imaging and SEM of perfusion decellularized cell sheets...... 64 Figure 4.1: Decellularized constructs fabrication for utilization in a prepared periodontal defect...... 79 Figure 4.2: Surgical defect preparation with decellularized construct utilization ...... 80 Figure 4.3: Graph showing different gene expressions in HPDL and human placental mesenchymal cells after seeding on decellularized matrix compared to PCL scaffold as a control at three time points...... 85

Figure 4.4: Stro-1 expression in human placental stem cells on PCL scaffold and decellularized periodontal ligament cell sheet 3, 7 and 14 days’ post seeding...... 88 Figure 4.5: micro CT 3D reconstructed image data showing created mandibular defects in rat mandibles treated with PCL scaffold, decellularized construct and fresh cell sheet with different degrees of bone fill at margins of the defects...... 91 Figure 4.6: Histological H&E stained sections of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery...... 92 Figure 4.7: Histological Mason Trichome stained sections of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery...... 93 Figure 4.8: Osteocalcin (OCN) immunohistochemistry staining of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery...... 96 Figure 4.9: Osteopontin (OPN) immunohistochemistry staining of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery ...... 98

Abstract

The periodontium is a complex structure that is comprised of soft tissue components the gingiva and periodontal ligament, as well as hard tissue components in the form of alveolar bone and cementum. Human periodontal ligament cells (HPDLCs) and their extracellular matrix are regarded as essential components to achieve successful periodontal regeneration when treating periodontal lesions.

In the present study, cell sheet technology was utilized to fabricate periodontal ligament cell sheets. These cell sheets were subsequently decellularized to isolate and preserve the extracellular component structural and functional characteristics, and were evaluated in vitro and in vivo. In all experiments that were undertaken in this study, melt electro-spun polycaprolactone (PCL) scaffolds were used as a carrier for the cell sheets, in order to support their fragile nature during the processes of decellularization. The combined cell sheet-PCL scaffold structure is referred to as a decellularized cell sheet construct.

In the first part of the study (chapter 2), human periodontal ligament cell sheets were decellularized using a simplified in-house perfusion system. Decellularized periodontal ligament sheets were evaluated for DNA elimination efficiency and collagen content preservation. Basic fibroblast, vascular endothelial and hepatocyte growth factors (b-FGF, VEGF and HGF) were quantified before and after decellularization. Decellularized matrices were also characterized using both confocal imaging and scanning electron microscopy (SEM). The recellularization potential of the decellularized sheets was assessed using allogenic periodontal ligament fibroblasts, where decellularized sheets were compared to a commercially available collagen membrane for the efficiency of recellularization. Results showed efficient DNA removal of 91%, with preservation of the collagen component. 10% of growth factors were retained in the decellularized sheets. Recellularization potential was evident over 21 days with cell proliferation and formation of new extracellular matrix on top of the decellularized sheets.

The second part of the study (chapter 3) aimed to compare different decellularization methods using both static and perfusion methods. A more optimized perfusion system was designed to enhance the efficiency of the decellularization and upscaling the number of sheets that could be decellularized without compromising the

DNA elimination. This study concluded that the combination of NH4OH and Triton X-100 together with a DNase treatment step was the most efficient method for DNA removal, preservation of extracellular matrix integrity and growth factors retention, irrespective of whether a static or perfusion approach was used.

The final part of the study (chapter 4) aimed to assess bone and PDL marker gene expression by HPDLCs and human placental stem cells (HPSCs) seeded on the decellularized sheets. Furthermore, the expression of the differentiation marker STRO-1 was evaluated using immunocytochemistry. Finally, the biocompatibility and regeneration potential of decellularized periodontal ligament cell sheet constructs was evaluated and compared to empty PCL scaffolds and fresh cell constructs in vivo using a rat mandibular periodontal defect model. The results showed upregulation of Osteocalcin (OCN), Osteopontin (OPN), bone morphogenetic protein-2 (BMP-2) and collagen type 1 A1 & 2 (COL1A1 &A2) in HPDLCs seeded on the decellularized sheets. STRO-1 expression was shown to decrease over 14 days of seeding on the decellularized sheets compared to empty controls. Biocompatibility and tissue integration was evident in vivo in periodontal defects treated with decellularized sheet constructs similar to the results obtained with fresh cell sheet constructs. Further, OCN and OPN immunostaining showed positive expression inside the defect area.

In conclusion, decellularized periodontal ligament cell sheets preserved their structural integrity, collagen content, and retained substantial amounts of growth factors in their extracellular matrix. Furthermore, they have the potential of recellularization as shown in both in vitro and in vivo experiments and have the potential to induce cells to express bone and collagen genes essential for the regenerative process.

XII

Acknowledgements

The joy of completion can only be felt when looking back at the journey past and remembering all the loving family and friends whom help and support eased this long but fulfilling journey.

First and foremost I would like to express my heartfelt deep gratitude to my principal supervisor, Professor Sašo Ivanovski, for his outstanding, limitless support and encouragement throughout my journey in my PhD study. His continuous guidance, sparing no time, effort or help were very much appreciated. Also, I would like to express my appreciation and gratitude to all my co-supervisors; Professor Dietmar Hutmacher, Dr. Cedryck Vaquette, Dr. Saeed Hashimi and Dr. Stephen Hamlet for their huge efforts and priceless help and guidance. Thank you all for bringing this PhD work to the light.

All thanks to my beloved country, The Arab Republic of Egypt who sponsored me, spared no expenses toward my education, and helped me to fulfil my journey achieving my PhD degree. I also thank Griffith University for funding my research, giving me amazing help and support and being the best place for me to undertake my PhD.

A very special thank you goes to my beloved wife, Yomna Bazaraa, who patiently supported me during my PhD, selflessly raised our daughters and always been the one to ease the hardships. As for my sweetest daughters, Mariam and Leena, who were the real motivation for me to keep trying and working harder to secure them a better life. My family, friends and colleagues thank you for your continuous help and support.

XIII

Publications and successful grants arising from this thesis

1. Farag A, Vaquette C, Theodoropoulos C, Hamlet SM, Hutmacher DW, Ivanovski S (2014). Decellularized periodontal ligament cell sheets with recellularization potential. Journal of dental research 93(12):1313-1319.

2. Farag A, Hashimi SM, Vaquette C, Volpato FZ, Hutmacher DW, Ivanovski S. Assessment of static and perfusion methods for decellularization of PCL membrane supported fibrous cell sheet constructs. Tissue engineering part C: Methods. Submitted.

3. Farag A, Hashimi SM, Vaquette C, Hutmacher DW, Ivanovski S. Decellularized periodontal ligament cell sheets for periodontal regeneration. Draft.

4. Periodontal Regeneration with Biphasic Scaffold and Cell Sheets. IADR Asia/Pacific Region (APR) Regional Meeting and Co-Annual Scientific Meeting of IADR Divisions 2013.

5. Vaquette C, Farag A, Saifzadeh S, Hutmacher D, Ivanovski S. (2015) Periodontal Regeneration with Biphasic Scaffolds and Cell Sheets. Draft.

6. National Health and Medical Research Council (NHMRC) grant APP1086181 titled: Evaluation of tissue engineered decellularised biphasic constructs for periodontal regeneration of $560K value.

XIV

Chapter 1

Introduction and Literature Review

Introduction:

Periodontal disease, an inflammatory disease affecting the soft and hard tissues surrounding the teeth, is initiated by the host in response to bacterial biofilm on the tooth surfaces adjacent to the gingival margin. Chronic periodontitis is manifested clinically by apical migration of junctional epithelium and loss of attachment, the formation of periodontal pockets, gingival recession and root exposure, tooth hyper-mobility and if left untreated, subsequent loss of the tooth (Armitage et al., 2003). Successful treatment regimens are aimed at preventing disease progression. However, due to the complex structure of the periodontium, which is composed of both soft (gingiva and periodontal ligament), and hard tissues (cementum and alveolar bone), the ultimate treatment goal of complete regeneration of damaged periodontal tissue is difficult to achieve.

Many different treatment modalities for periodontal regeneration have been used with limited degrees of success. These include guided tissue regeneration using non-resorbable and bio-absorbable membranes with or without concomitant use of bone substitutes, and bioactive materials such as enamel matrix derivatives, growth factors and bone morphogenetic proteins. However none of the aforementioned treatment modalities has completely fulfilled the ultimate goal of complete structural and functional periodontal tissue regeneration.

1 Recently, the engineering of periodontal tissues has been investigated as a possible treatment modality to restore lost periodontal tissues. Periodontal tissue engineering is based upon the concept of growing periodontal cells in vitro before implanting them surgically into a defect site. A novel approach to tissue engineering using ‘cell sheet’ technology has been proposed whereby contiguous cell monolayers complete with extracellular matrix (cell sheets) could be produced from various cell types including periodontal ligament cells (Ishikawa et al., 2009). Periodontal ligament cell sheets were subsequently shown to promote periodontal tissue regeneration in vivo (Akizuki et al., 2005; Flores et al., 2008b).

In a previous study from our group, it was shown that periodontal ligament cell sheets prepared in vitro and delivered using a novel biphasic scaffold, were capable of simultaneous regeneration of bone and periodontal ligament (Vaquette et al., 2012). Furthermore, another study evaluated the influence of the cell type used to prepare the cell sheets on regenerative outcomes in a surgically created rat periodontal defect model (Dan et al., 2014). The results demonstrated the regenerative potential of periodontal ligament cell sheets, with new functional periodontal ligament, alveolar bone and new cementum formed within the defect (Dan et al., 2014).

While a tissue engineered cell sheet approach is very promising, there are several underlying limitations hindering this technology from being applicable to clinical practice. One such limitation is the reliance on the patient’s own cells, resulting in problems associated with reproducibility and safety. Another limitation is that cell sheet technology requires cell culture facilities, technical expertise and thus considerable expense to bring this technology into daily use in the clinic.

To overcome these limitations, the use of decellularized matrices is gaining attention for tissue engineering applications. Indeed, many studies have evaluated the effectiveness of decellularized matrices in tissue engineering and regeneration applications (Bloch et al., 2011; Dijkman et al., 2012; Dong et al., 2009; Fang et al., 2007; Flynn et al., 2006; Hoshiba et al., 2010; Schmidt et al., 2007; Shimizu et al., 2007; White et al., 2005; Yazdani et al., 2009). These studies have shown that the matrices retain most of their biological and mechanical properties after the process of decellularization and do not induce an immune response in vivo. Notably, commercially available products, such as Alloderm® and Mucograft®, have been

reported to achieve favorable outcomes in periodontal cosmetic surgery applications, achieving comparable results with the current gold standard of autogenous subepithelial connective tissue grafts (Awad et al., 2009; Chen et al., 2010; Cothren et al., 2004; Kish et al., 2005; Korn et al., 2009; Nevins et al., 2011; Rotundo and Pini- Prato, 2012; Salzberg et al., 2011; Sanz et al., 2009; Takahashi et al., 2011; Wainwright, 1995).

In addition to decellularized intact tissues and organs, cell derived tissue engineered decellularized matrices prepared in vitro have also been shown to retain their structural integrity and maintain their molecular functionality; and when used in vivo supported by a suitable scaffold, they enhance tissues regeneration such as bone and cartilage (Gawlitta et al., 2015; Papadimitropoulos et al., 2015; Sadr et al., 2012). However, the use of tissue engineered decellularized constructs has not been evaluated in the field of periodontal regeneration.

Literature Review

Structure of the periodontium:

The periodontium is a group of tissues with distinct structures that support and invest the tooth. It is composed of four different tissues; the cementum lining the tooth root, the periodontal ligament, alveolar bone that lines the socket, and the gingiva facing the tooth known as the dento-gingival junction. Healthy periodontium is characterized by the structural integrity and interaction between its different components (Nanci and Bosshardt, 2006b).

Cementum:

Cementum is avascular calcified connective tissue that covers the tooth root surfaces. It consists of calcified ground substance (organic matrix) together with embedded collagen fibers. These fibers are mainly type I (90%) although type III and VII are present in lesser amounts. Cementum is classified according to the presence or absence of cells (cementocytes), and also according to the origin of the collagen fibers

i.e. whether they are derived from the principle periodontal ligament fibers or intrinsic fibers synthesized by cementoblasts (Cho and Garant, 2000).

Acellular cementum (also called acellular extrinsic fiber cementum or primary cementum) covers the cervical and middle portions of the tooth root. It does not contain cells and is characterized by the large number of inserting principal periodontal ligament fibers known as “Sharpey’s fibers”. Thus acellular cementum plays an important role in tooth attachment (Bosshardt and Selvig, 1997), and its formation is essential for periodontal regeneration.

Cellular cementum is present on the apical third or sometimes the apical half of the tooth root. It is less mineralized than acellular cementum due to its rapid formation. It is characterized by the presence of cementocytes in lacunae embedded within the substance of cementum, and the intrinsic collagen fibers synthesized by the cementoblasts (Bosshardt, 2005). Cellular cementum is important for homeostasis and repair.

Periodontal ligament:

The periodontal ligament is a specialized connective tissue compartment that fills the gap between the surface cementum covering the tooth root and the alveolar bone lining the socket. Its thickness varies from 0.15 to 0.38 mm. The periodontal ligament supports the tooth in its socket, allows the tooth to withstand masticatory forces, provides a sensory function due the presence of proprioceptive receptors , and acts as a cell reservoir for the homeostasis of periodontal tissues (Berkovitz, 2004; Nanci and Bosshardt, 2006b).

The periodontal ligament extracellular matrix consists mainly of well- organized bundles of collagen fibers embedded in amorphous ground substance. The periodontal ligament fibroblast is the predominant cell type, and it is found attached and aligned in the same direction as the bundles of collagen fibers. The main function of periodontal ligament fibroblasts is the metabolism of extracellular matrix components, especially collagen, with fibroblasts having the capability to synthesize and degrade collagen simultaneously (Cho and Garant, 2000; Nanci and Bosshardt, 2006b). There is a heterogeneous population of fibroblasts in the periodontal

ligament, as exemplified by subpopulations exhibiting high alkaline phosphatase activity that are able to differentiate into bone forming cells and cementoblasts. Therefore, periodontal ligament fibroblasts are a key cell type in periodontal wound healing (Cho and Garant, 2000; Ivanovski et al., 2001; Murakami et al., 2003b).

Undifferentiated mesenchymal cells of the periodontal ligament, also known as periodontium derived stem cells (PDSCS) are another cell population that are of ecto-mesenchymal origin and have been characterized as a population of multipotent stem cells (Arnold et al., 2010; Seo et al., 2004). They play an important role in periodontal wound healing by facilitating the repair and regeneration of periodontal ligament, alveolar bone and cementum (Ivanovski et al., 2006; Trubiani et al., 2008). Interestingly, PDSCS could be isolated using neural stem cell specific biomarkers as nestin and Sox2, then cultured in a serum-free medium containing basic fibroblast growth factor (bFGF). These cells were found to form spheres, previously described as neurospheres (Widera et al., 2007), after 8-10 days of culture and when cultured in osteogenic medium, the cells were capable of differentiating into osteogenic precursor cells, which produce extracellular matrix containing osteopontin (Arnold et al., 2010).

Periodontal ligament fibers can be classified into collagenous and non- collagenous (elastic) fibers. Collagen (predominantly type I, III an XII) fibrils are organized into bundles termed principal periodontal ligament fibers. The terminals of theses principal fibers are embedded into alveolar bone on one end and into cementum on the other, with the embedded part known as Sharpey’s fibers. Sharpey’s fibers embedded into acellular cementum are fully mineralized however those embedded in cellular cementum or alveolar bone are only partially mineralized.

Alveolar bone:

Alveolar bone can be sub-classified into alveolar bone proper and supporting alveolar bone. Alveolar bone proper is the bone forming the wall of the tooth socket, consisting of bundle bone supported with haversian and lamellated bone, with a thickness of 0.1 to 0.4 mm. It contains Sharpey’s fibers embedded within its matrix. The supporting alveolar bone consists of cortical plates of compact bone on the buccal, lingual and palatal aspects, with spongy bone filling the space between the

5 inner and outer plates of cortical bone and the alveolar bone proper (Cho and Garant, 2000).

Gingiva:

The gingiva is a soft tissue with epithelial and connective tissue compartments. It is the part of the oral mucosa that covers the cervical aspect of the teeth and the alveolar bone. The gingival epithelium is divided into three parts: gingival, sulcular and junctional epithelium, the turnover rate being fastest in junctional epithelium and slowest in gingival epithelium. In normal healthy periodontium the junctional epithelium ends at the cemento-enamel junction and is attached to the tooth surface by means of hemidesmosomes and internal basal lamina, which consists of laminin and proteoglycans (Cho and Garant, 2000; Nanci and Bosshardt, 2006b).

The gingival connective tissue compartment contains a highly vascular plexus, as well as bundles of collagen fibers that run from the cervical cementum and insert into the gingival connective tissue, its function being to attach the gingiva to the tooth surface. The integrity of these fibers together with the junctional epithelium is essential for protection of the underlying periodontal ligament, cementum and alveolar bone from bacteria and the oral environment (Nanci and Bosshardt, 2006b).

Effect of periodontal disease on periodontal tissues:

Junctional epithelium and gingival connective tissues:

Periodontal disease starts is initiated by the colonization of periodontal pathogens on the tooth surface and the formation of a supragingival plaque biofilm. In susceptible patients, a host response is initiated by bacterial surface components such as lipoteichoic acids and lipopolysaccharides that results in the loss of integrity of the junctional epithelium and gingival connective tissue, allowing subgingival colonization by periodontal pathogens (Pollanen et al., 2000). Periodontal pathogens such as Porphyromonas gingivalis which is considered to be one of the major

6 etiologic factors of periodontal disease, release potent gingipains and matrix metalloproteinases (MMPs) that may also degrade the connective tissue matrix (Groeger et al., 2010; Kadowaki and Yamamoto, 2003; Nanci and Bosshardt, 2006b; NM et al., 2003).

Periodontal ligament and alveolar bone:

Periodontal ligament fibers and extracellular matrix are not only degraded by bacterial gingipains and matrix metalloproteinases, but also by host produced pro- inflammatory cytokines, mainly IL-1, IL-6 and tumor necrosis factor alpha (TNF-α), secreted during the immune response to periodontal pathogens (Nanci and Bosshardt, 2006b). After triggering the host inflammatory process, neutrophils release large amounts of elastase and matrix metalloproteinases (MMPs), especially collagenases, causing progressive destruction of periodontal tissues and alveolar bone matrix (Potempa et al., 2000).

Alveolar bone undergoes resorption as a result of the stimulation of osteoclastic activity, which is triggered by host-produced inflammatory cytokines such as IL-1, IL-6, IL-11, IL-17, PGE2 and TNF-α (Nanci and Bosshardt, 2006b).

Periodontal regeneration:

For decades periodontal therapy has focused on stopping the progression of the periodontal disease. However, research is now also directed towards restoring the lost structure and function of the periodontal tissues.

Some common terms widely used to describe the histological outcomes of different periodontal treatment modalities include:

• Periodontal repair: It is a form of healing that results in formation of a long junctional epithelial attachment along the root surface that was previously denuded by the periodontal disease and exposed to the periodontal pocket environment and to periodontal pathogens. Histologically it is characterized by a thin epithelium extending apically interposed between the root surface

and the gingival connective tissue. This form of healing is usually associated with minimal or no actual regeneration of lost periodontal tissues.

• New attachment: Also known as connective tissue repair. It is a treatment outcome that is usually represented by periodontal ligament collagen fibers oriented parallel or perpendicular to root surface that was previously exposed to periodontal disease or deprived of , together with formation of new junctional epithelium attached to tooth established at its original location before the damage caused by the periodontal disease.

• Periodontal regeneration: Restoration of both the structure and function of periodontal tissues destroyed by the periodontal disease. This is achieved by formation of new (de novo) periodontal ligament attachment inserted into newly formed cementum on root surface that was previously denuded by periodontal disease with simultaneous regeneration of alveolar bone and gingiva that were damaged during periods of disease activity.

Periodontal regenerative therapies:

In this review different regenerative techniques in periodontal therapy will be evaluated with respect to their clinical and histological results. The potential of any new biomaterial or therapy for periodontal regeneration is often first evaluated by histological analysis in a suitable animal model (Kim et al., 2011; Polimeni et al., 2006).

While most conventional periodontal treatments may have successful clinical outcomes in term of clinical attachment gain and reduction of periodontal pocket probing depth, they are generally characterized by the formation of long junctional epithelial attachment formation between the gingival connective tissue and the root surface (Ivanovski, 2009; Kim et al., 2011).

Histological analysis of healing following conventional non regenerative surgical procedures in surgically created class III defects in animals has shown that most of the defects were filled with gingival connective tissue and formation of long

junctional epithelium. While new cementum and alveolar bone have been detected, this was limited to the most apical part of the defect. Root resorption and/or ankyloses were generally not observed (Deliberador et al., 2006; Kim et al., 2011; Lindhe et al., 1995).

Guided tissue regeneration (GTR):

GTR is a technique used in periodontal therapy to prevent epithelial cells from migrating along the tooth surface of the periodontal pocket. It is based on the premise that periodontal ligament cells are the only type of cells capable of regenerating the periodontal tissues and the attachment apparatus of the tooth (Gottlow et al., 1986; Nyman et al., 1982).

GTR is a surgical procedure that involves the placement of a barrier membrane that excludes the epithelium and gingival connective tissue from the periodontal defect and prevents epithelium migration along the debrided root surface. This allows the periodontal ligament cells to selectively repopulate the periodontal defect and form new attachment, new cementum and new bone (Dahlin et al., 1988; Nyman et al., 1987).

GTR has been shown to be suitable for treating intrabony and Class II mandibular furcation defects (Cortellini and Tonetti, 2000; Sanz and Giovannoli, 2000). Several histological studies performed in animal models (Dahlin et al., 1988; Gottlow et al., 1990) and on human biopsies (Gottlow et al., 1986; Nyman et al., 1982) showed GTR to have regenerative potential capable of achieving new attachment with new cementum and bone formation.

There are many different types of barrier membranes that have been used for GTR, which have been broadly classified as either non-resorbable or resorbable.

Non-resorbable Barrier membranes:

Expanded polytetrafloroethylene (e-PTFE) has been the most commonly used non-resorbable barrier (Gore-Tex, W.L. Gore & associates, Inc, Newark, DE) (Bashutski and Wang, 2009). Some studies have shown that intrabony defects treated

9 using, e-PTFE membranes could result in 3.0-5.0 mm bone fill with 4.0 -7.0 mm CAL gain with or without using bone graft/bone substitutes (Bashutski and Wang, 2009; Cortellini and Tonetti, 2005; Dori et al., 2007b; Dori et al., 2008a).

Cortellini & Tonetti demonstrated that treatment of intrabony defects using non-resorbable barrier membranes resulted in clinical attachment level (CAL) gain and reduction of probing pocket depth (PD) compared to conventional access flap surgery (Cortellini and Tonetti, 2000).

Schallhorn & McClain demonstrated that it is possible to achieve significant improvement in both CAL gain and reduction of PD by combining barrier membrane with bone/ bone substitutes ((McClain and Schallhorn, 1993), however Cortellini & Tonetti reported that bone substitutes failed to demonstrate an additive effect compared to barrier membranes used alone in the treatment of deep intrabony defects (Cortellini and Tonetti, 2000).

One of the most common complications associated with the use of non- resorbable membranes in GTR is barrier membrane exposure (Becker et al., 1988; Cortellini et al., 1990; Cortellini et al., 1993; De Sanctis et al., 1996b; Murphy, 1995). Barrier membranes exposed after surgical procedures have been found to be contaminated by bacteria, which negatively affects the clinical outcome of the GTR surgical procedures. This was demonstrated by a significant decrease in CAL gain (De Sanctis et al., 1996a; b; Nowzari et al., 1995; Selvig et al., 1992). However, membrane exposure may be significantly reduced by the use of advanced surgical procedures, such as the modified papilla preservation technique ((Checchi et al., 2009; Cortellini et al., 1995; Cortellini et al., 1996).

Bioresorbable barrier membranes:

The main clinical difference between bioresorbable and non-resorbable barrier membranes is that the later requires a second surgical procedure after 4 to 6 weeks for membrane removal. Second generation bioresorbable barrier membranes have subsequently been developed which are readily absorbed and do not require a re- entry surgical procedure, and have superseded non-resorbable membranes for most clinical applications.

There are various types of materials used to manufacture bioresorbable membranes, such as collagen, polyglactin 910, polylactic acid (PLA), polylactic acid- polyglycolic acid co-polymer (PLA-PGA), and acellular dermal matrix allograft (ADM) (Bashutski and Wang, 2009).

The most commonly used bioresorbable membranes are collagen membranes, which can have various degrees of cross-linking and hence have different resorption rates (Bashutski and Wang, 2009). Collagen membranes are widely used for their ease of manipulation, handling and decreased post-operative complications. For example BioGide (Geistlich, Wolhusen, Switzerland), which is of porcine dermis origin consists of non-cross-linked bilayer collagen type I & II that is resorbed within 2-4 weeks. BioMend (Sulzer Calcitek, Carlsbad, CA) is of bovine tendon origin and consists of cross-linked collagen type I, with a resorption rate that ranges from 6 to 8 weeks (Bashutski and Wang, 2009; Bunyaratavej and Wang, 2001).

Bioresorbable membranes have been shown to be effective in periodontal regeneration in numerous studies with significant clinical improvement in the amount of bone fill, CAL gain and reduction of probing PD, in both deep intrabony (Becker et al., 1988; Cortellini and Tonetti, 2000; Stavropoulos and Karring, 2004) and furcation defects (McClain and Schallhorn, 2000; Sanz and Giovannoli, 2000).

Some studies have shown that collagen membranes may also have a beneficial effect in wound healing by promoting platelet aggregation and fibroblast migration, hence enhancing hemostasis and accelerating the healing process (Bashutski and Wang, 2009; Steinberg et al., 1986).

Although GTR using both resorbable and non-resorbable membranes have been shown to promote periodontal regeneration in favorable defects (intrabony, Class 2 furcation), clinical outcomes can be considered unpredictable and technique sensitive (Ivanovski, 2009).

Bone Grafting:

The aim of bone grafting in periodontal bony defects is to promote bone formation, with the expectations that this would result in new periodontal attachment along the root surface (Ivanovski, 2009).

Autogenous bone graft:

Considered as the gold standard for bone replacement grafting due to its osteogenicity and relatively low cost, autogenous grafts can be harvested either intraorally from the extraction socket, edentulous ridge, ramus, symphysis and tuberosity or; extraorally from the iliac crest and tibia. The main disadvantage of autogenous bone grafts is the need for a second surgical site and the associated increased patient morbidity (Bashutski and Wang, 2009).

It has been shown that extraoral bone grafts is a successful method for the treatment of periodontal bony defects, with an average of 4mm in bone height gain when treating crater defects (Schallhorn, 1968). However other groups have reported root resorption after using autogenous extraoral bone grafts in periodontal defects (Dragoo and Sullivan, 1973).

Intraoral autogenous bone grafts have been widely used in the treatment of periodontal bony defects. Some studies have provided evidence of vertical bone gain of an average 3.5 mm (Hiatt and Schallhorn, 1973), while others have reported new periodontal ligament and new cementum formation with new attachment (Mellonig and Bowers, 1990).

Allografts:

To avoid a second surgical operation to harvest a bone graft, allografts have been used extensively with results comparable to autogenous bone. The most commonly used allografts are freeze-dried bone allograft (FDBA) and demineralized freeze-dried bone allograft (DFDBA). Their main disadvantages are their high cost

12 and that this type of graft may trigger a foreign body immune response (Bashutski and Wang, 2009).

In 1965 Urist demonstrated that decalcification of the bone graft induces an osteoinductive effect when placed into bony defects, stimulating progenitor bone cells to differentiate into bone forming cells that form new bone. This inductive effect has been shown to be due to bone morphogenetic proteins (BMPs) that are exposed after decalcification (Urist, 1965). However decalcification of allograft bone makes it degrade and resorb at a higher rate when compared to FDBA, decreasing its efficiency as a scaffold (Bashutski and Wang, 2009).

Clinical evaluations of FDBA have reported an average bone fill of 1.3 - 2.6 mm in the treatment of intrabony periodontal defects (Altiere et al., 1979; Bashutski and Wang, 2009; Blumenthal and Steinberg, 1990; Mellonig, 1991). Evaluation of DFDBA showed bone fill similar to that of FDBA with an average bone fill of 1.7 - 2.9 mm (Mellonig, 1984; Pearson et al., 1981).

In a systematic review based on 49 controlled clinical studies evaluating the clinical outcome of using bone replacement grafts (BRG) in the treatment of intrabony defects, and 17 controlled clinical studies evaluating BRG in the treatment of furcation defects (Reynolds et al., 2003), it was concluded that FDBA showed an improvement in clinical outcomes in terms of clinical attachment gain and reduction in periodontal pocket depth in the treatment of intra bony defects (Barnett et al., 1989; Mellonig et al., 1976; Sanders et al., 1983).

In a retrospective study comparing the use of DFDBA to freeze-dried bone allograft (FDBA) combined with enamel matrix derivative (EMD) for the treatment of advanced osseous defects, it was concluded that FDBA combined with EMD resulted in superior clinical outcomes in term of clinical attachment level gain compared to that obtained with DFDBA (Rosen and Reynolds, 2002). Piattelli et al (1996) also suggested that FDBA may have better osteoinductive potential than DFDBA (Piattelli et al., 1996).

Xenografts:

Xenografts have been widely used as a bony substitute in the treatment of intrabony defects. Usually harvested from bovine or porcine sources, they are processed in order to remove all organic constituents that might elicit a human immunogenic response, leaving only an anorganic matrix that acts as a scaffold which usually resorbs very slowly (Bashutski and Wang, 2009).

Many studies have evaluated xenografts in the treatment of intrabony and furcation periodontal defects and reported improvements in clinical outcomes in term of reduction of PPD and gain in CAL (Houser et al., 2001; Sculean et al., 2004a).

Xenografts have also been used in combination with other regenerative techniques in the treatment of periodontal bony defects (Sculean et al., 2004b; Yilmaz et al., 2009; Zafiropoulos et al., 2007). Sculean et al (2002) clinically evaluated bovine derived xenografts (BDX) combined with enamel matrix derivatives (EMD) compared to BDX alone in the treatment of 12 intrabony defects. Both techniques resulted in clinical improvement in pocket probing depth and gain in clinical attachment levels. Furthermore, all defects showed hard tissue fill with no statistical significant difference between the two groups (Sculean et al., 2002).

Alloplasts:

Alloplasts are synthetic materials utilized mainly as space maintaining bone fillers. Hydroxyapatite and beta-tricalcium phosphate are the most widely used alloplasts. These compounds act as osteoconductive rather than osteoinductive materials and usually result in fibrous encapsulation without any signs of periodontal tissue regeneration (Ivanovski, 2009; Reynolds et al., 2003). Therefore periodontal tissue regeneration with alloplastic bone substitutes is very unpredictable (Bashutski and Wang, 2009).

Bioactive materials & signaling molecules:

Enamel matrix derivative (EMD):

EMD is a biologically active regenerative material that has been used widely in regenerative periodontal therapies. It is harvested from developing porcine teeth and has been reported to induce periodontal regeneration, as demonstrated by new cementum, periodontal ligament attachment and bone formation in intrabony defects in humans (Mellonig, 1999; Sculean et al., 2000; Yukna and Mellonig, 2000).

EMD contains a mixture of low molecular weight proteins that belong to the amelogenin family. These proteins promote fibroblast like cells to repopulate denuded root surfaces after the application of EMD (Gestrelius et al., 1997), and stimulate cell growth and differentiation of mesenchymal cells into osteoblasts (Hagewald et al., 2004; He et al., 2004; Kao et al., 2009).

EMD has been shown to up regulate transforming growth factor-beta (TGF-β) and bone morphogenetic protein expression in osteoblasts It also stimulates new blood vessel formation due to its chemotactic and proliferative effect on endothelial cells which is an essential stage in wound healing and tissue regeneration (Schlueter et al., 2007).

Many studies have shown improved clinical outcomes using EMD in the treatment of intrabony defects, in term of reduction of periodontal probing depth (PPD) and gain in clinical attachment level (CAL) (Heard et al., 2000; Manor, 2000). Rasperini et al. (2000) demonstrated that EMD improved clinical results with a mean of 2mm gain in CAL and an average of 3mm increase in keratinized tissue height. Histologically the authors showed that there was new cementum formation with new connective tissue fiber attachment and evidence of new woven bone formation (Rasperini et al., 2000). Froum et al. (2001) compared the treatment of intrabony defects treated with EMD open flap debridement only and showed that the average percent of defect bone fill was more than three times greater in the EMD treated group. These authors concluded that treatment of periodontal intrabony defects with EMD is clinically superior to open flap debridement in every parameter evaluated,

and reported that re-entry data demonstrated that the percentage of bone fill is comparable to other studies that utilized bony grafts or GTR(Froum et al., 2001).

EMD has also been assessed in combination with other regenerative procedures. Some studies combined EMD with bovine xenografts (Scheyer et al., 2002; Sculean et al., 2002), and reported that this combination may improve bone fill, although no statistically significant difference was found compared to sites treated with EMD alone. Other studies compared EMD combined with bioactive glass to EMD alone and showed that there were no statistically significant differences between the two groups in terms of reduction of pocket depth or gain in CAL in short or long term follow up periods (Sculean et al., 2005; Sculean et al., 2007a). According to recent systematic reviews evaluating the combination of EMD and GTR in the treatment of intrabony defects, it was concluded that this combination does not result in any additional improvement in clinical outcomes compared to the utilization of EMD alone (Rathe et al., 2009; Sculean et al., 2007b; Trombelli and Farina, 2008).

Platelet Rich Plasma (PRP):

Platelet rich plasma is defined as an autologous concentrate of platelets in a small volume of plasma and is considered to be a rich source of autologous growth factors (Marx, 2004) such as platelet derived growth factor (PDGF), transforming growth factor beta (TFG-β), vascular endothelial growth factor (VEGF) and epithelial growth factor (EGF) (Rodrigues et al., 2012). These growth factors play an important role in periodontal wound healing and PRP has shown to be effective in promoting periodontal regeneration in several studies.

Okuda et al. (2005) compared the combination of PRP with biodegradable porous hydroxyapatite (HA) to HA alone in the treatment of intrabony defects over 12 months and found that the combination of PRP with HA showed a statistical significant improvement in periodontal probing depth (PPD) reduction and clinical attachment gain(Okuda et al., 2005). Another study evaluating PRP in combination with bovine porous bone material graft (BPBM) showed superior reduction in PPD, gain in CAL and bony defect fill after 12 months compared to BPBM alone (Ouyang and Qiao, 2006).

Demir et al. (2007) compared the effectiveness of PRP combined with bioactive glass (BG) to BG alone in the treatment of intrabony defects after 9 months. They showed that there were no statistically significant differences between the two groups in term of reduction of PPD, gain in CAL and defect bone fill (Demir et al., 2007). Another study which evaluated PRP usage combined with natural bone mineral and GTR showed that both treatments resulted in significant improvement in reduction of PPD and gain in CAL (Dori et al., 2007a). Markou et al. (2009) showed that addition DFDBA to PRP did not significantly improve the treatment results when compared to PRP alone, with both groups demonstratinga reduction of pocket depths and gain in clinical attachment levels (Markou et al., 2009).

Platelet derived growth factor (PDGF):

Growth factors are proteins that are secreted by various host cells to regulate a wide variety of functions, such as cell chemotaxis, cell growth, proliferation, cell differentiation and extracellular matrix formation(Kao et al., 2009). PDGF has a potent effect in accelerating wound healing due to its chemotactic and mitotic effect on mesenchymal stem cells. Recombinant forms of PDGF have recently been utilized in periodontal regenerative therapies in the treatment of intrabony and furcation defects(Kao et al., 2009). PDGF and Insulin growth factor type 1 (IGF-1) have been shown to promote periodontal regeneration when topically applied to periodontaly diseased root surfaces in animal studies with histology revealing new cementum, alveolar bone and periodontal attachment (Kao et al., 2009; Murakami, 2011).

Recombinant human platelet derived growth factor combined to recombinant human insulin growth factor were utilized in a human clinical study to evaluate their regenerative potential. Two different doses were used, a low dose of 50 µg/ml and a high dose of 150 µg/ml. Sites receiving high dose growth factors showed bone fill on average of 43.2% compared to 18.5% in control sites, and 2.08 mm vertical new bone height (high dose) compared to 0.75 mm in control sites (Nevins et al., 2005).

The combination of rh-PDGF with tri-calcium phosphate (TCP) has been available commercially as GEM-21®. This product has been shown to improve

clinical outcomes in terms of CAL gain, reduction of PPD and radiographic bone fill (Kao et al., 2009; McGuire et al., 2006; Nevins et al., 2005).

Fibroblast growth factor (FGF):

Fibroblast growth factor is a polypeptide that consists of approximately 150- 200 amino acid residues. FGF stimulates undifferentiated mesenchymal cells to differentiate and proliferate, and can also induce angiogenesis (Murakami, 2011). Murakami et al. (2003) investigated the effect of topical application of recombinant human basic fibroblast growth factor on periodontal regeneration in furcation defects in beagle dogs and showed that sites treated with rh FGF-2 had histologically verified evidence of periodontal regeneration, with new cementum, new alveolar bone and new Sharpey’s fibers with functional new periodontal ligament (Murakami et al., 2003a).

However in a large randomized controlled double blind human clinical study no statistically significant differences were noticed in CAL gain and alveolar bone gain in FGF-2 treated subjects. There was however a significant difference in alveolar bone height between the two groups after 36 weeks suggesting FGF-2 could have a positive long term effect in stimulating regeneration in intrabony periodontal defects(Kitamura et al., 2008).

Bone morphogenetic proteins (BMPs):

Bone morphogenetic proteins are glycoproteins and members of the transforming growth factor-beta superfamily. These glycoproteins are known to stimulate stem cells to differentiate into osteoblasts and chondroblasts. Most of the research in periodontology has focused on three BMPs for their capability to enhance periodontal regeneration, namely BMP-2, BMP-3 and BMP-7 (Kao et al., 2009). Demineralized freeze dried bone allograft preparations contain BMPs that can enhance periodontal and bone regeneration. However it has been shown that the concentration in these bone preparations are low and variable, and hence clinically utilized BMPs are currently prepared using recombinant DNA technology (Schwartz et al., 1996; Schwartz et al., 1998). Recombinant human BMP-2 (rhBMP-2) has been

shown to enhance new cementum, alveolar bone and periodontal ligament formation, however in this animal study the authors reported areas of root resorption and ankyloses (Sigurdsson et al., 1995). Compared to rhBMP-2, rhBMP-7 did not show signs of ankyloses or root resorption (Kao et al., 2009).

Tissue engineering of periodontal ligament:

Tissue engineering is a concept based on the utilization of patients’ own cells to regenerate lost or damaged tissues after seeding the cells on an appropriate synthetic scaffold (Langer and Vacanti, 1993). Tissue engineering relies on three main factors: cells with regenerative potential, signaling molecules as instructive cues (to enhance proliferation and differentiation of progenitor cells) and finally a carrier or scaffold (Lin et al., 2009).

Progenitor cells:

In order to achieve periodontal regeneration using tissue engineering, appropriate progenitor cells need to be used. Stem cells have gained widespread attention for their capability for self-renewal and their ability to differentiate into multiple cell types (Rahaman and Mao, 2005). Stem cells can be of either embryonic or adult origin. Adult stem cells are abundant in bone marrow, adipose tissues (Zuk et al., 2001), and also available in periodontal ligament and pulp tissues (Miura et al., 2003).

Since stem cells have been shown to have the capability to differentiate into a variety of cell lineages under specific conditions, it has been proposed that they could be used to restore complex structures, such as periodontal tissues (Rahaman and Mao, 2005). Indeed, progenitor cells with the potential to differentiate into osteoblasts and cementoblasts have been isolated from the periodontal ligament (Trubiani et al., 2008). Further, progenitor cells with the potential to differentiation into other cell lineages have been identified in gingival tissues (Tomar et al., 2010).

Scaffolds used for cell delivery:

There are different scaffold materials that can be used as a carrier for cells in tissue engineering. Generally they can be divided into natural (as collagen, gelatin, chitosan etc.) and synthetic scaffolds (β-tricalcium phosphate (β-TCP), polyglycolic acid (PGA) polycaprolactone (PCL) etc.), and can also be classified as resorbable and non resorbable (Bartold et al., 2006). The rationale for using scaffolds in tissue engineering is to provide a framework to enhance extracellular matrices deposition (Rahaman and Mao, 2005).

Polycaprolactone (PCL) was one of the first synthetic polymers that was utilized as a biodegradable material. It has a hydrophobic semi-crystalline nature (Woodruff and Hutmacher, 2010). When implanted in vivo, PCL was found to have a slow degradation rate ranging between 2 and 4 years (Pitt et al., 1981). In contrast to other commonly used biomaterial polymers, such as poly-glycolic acid (PGA) and poly-lactic acid (PLA), The slow degradation results in the formation of low concentrations of potentially cytotoxic by-products, and hence PCL is widely recognized of being a very biocompatible material. Indeed, it has been reported that PCL scaffolds are not cytotoxic towards human osteoprogenitor cells in vitro (Schappacher et al., 2010).

Several studies have evaluated both the short- and long-term in vivo biocompatibility of PCL scaffolds using animal models, and reported that PCL and PCL composites have no adverse effects at a variety of healing timepoints, ranging from 15 weeks to 2 years (Lam et al., 2007; Sawyer et al., 2009; Woodruff and Hutmacher, 2010).

PCL scaffolds have been widely utilized for a variety of tissue engineering applications, such as drug delivery (Yoon and Kim, 2011), cell delivery (Dan et al., 2014) and tissue engineering. Notably, PCL scaffolds have been widely used for the fabrication of biomimetic constructs for bone regeneration (Ko et al., 2015; Remya et al., 2013; Shao et al., 2015; Tambralli et al., 2009).

In this thesis, all of the experiments relied on the utilization of PCL scaffolds and membranes, whereby melt electrospun scaffolds were used to support cell sheets during harvesting and decellularization processes.

Cell sheet technology:

The harvesting of cell sheets using thermo-responsive cell culture plates is a tissue engineering technology that was developed by Okano and co-workers, initially utilizing corneal epithelial cell sheets to make transplantable corneas (Kobayashi et al., 2013; Nakajima et al., 2012; Nakajima et al., 2015). This group also used the same concept to develop tissue engineered myocardial tissues using multi-layered cardiomyocyte cell sheets (Zhou et al., 2007). In 2005, Okano’s and Ishikawa’s groups collaborated on a study that used cell sheet technology to assess the capability of periodontal ligament cells to facilitate periodontal regeneration. The results showed successful periodontal regeneration in terms of new cementum, new bone and new periodontal ligament in 60% of the test group (Akizuki et al., 2005).

Flores et al. (2008) further assessed the regenerative potential of human periodontal ligament cell sheets in surgically created fenestration periodontal defects in athymic rats. The authors reported that growing human periodontal ligament cells in osteogenic differentiation media prior to implantation led to the increased expression of the bone markers osteopontin (OPN) and bone sialoprotein (BSP). Histological analysis later also confirmed that the cell sheets were capable of regenerating cementum, bone and new periodontal ligament fibers that were perpendicularly inserted into the newly formed cementum (Flores et al., 2008b). These studies confirmed that periodontal ligament cell sheets contain a diverse and heterogeneous subpopulation of cells that are capable of differentiating into PDL, cementum and bone tissues.

Because of the delicate and fragile nature of the cell sheet, this technology also requires a supportive membrane to facilitate its manipulation and implantation into the defect site. In a recent study, a novel polycaprolactone biphasic scaffold was used for simultaneous bone and periodontal ligament regeneration (Vaquette et al., 2012). This scaffold was composed of a Fused Deposition Modeling (FDM) scaffold for the bone compartment and a flexible electrospun membrane for the periodontal ligament compartment. Cell sheets were prepared from sheep PDL for the periodontal ligament compartment and osteoblasts were seeded in the bone compartment. This complex construct was secured to a dentine slice and then implanted subcutaneously into athymic nude rats. The results showed new mineralized cementum like tissue on the

dentin along with mineralized tissue in the bone compartment, with evidence of new PDL like fibers inserted into the newly formed cementum (Vaquette et al., 2012).

This promising tissue engineering approach (cell sheets combined with novel biphasic polycaprolactone scaffolds) was further investigated by our group, whereby cells from different periodontal sources (gingiva, PDL, alveolar bone) were investigated in vitro and in vivo for their periodontal regenerative potential. Micro computed topography, H&E staining and immunohistochemistry were used to evaluate periodontal healing in surgically prepared periodontal defects in a rat model. The results showed that periodontal ligament cell sheets were superior to alveolar bone and gingival cell sheets in terms of cementum, bone and functional periodontal ligament formation (Dan et al., 2014). Based on these findings, periodontal ligament cell sheets were used in the studies described in this thesis.

Tissue engineered decellularized matrices and periodontal regeneration:

The concept of using decellularized matrices as a biologic scaffold has gained wide attention lately. The rationale of using such a method is to produce three dimensional scaffolds that mimic natural tissue composition, microstructure, biological and mechanical properties, hence enhancing the recruitment of the host cells into these scaffolds, allowing them to differentiate and secrete new extracellular matrix (ECM) resulting in regeneration of damaged tissues.

There are several different ways to obtain ECM scaffolds, which can be broadly divided into approaches using decellularized harvested intact tissues and organs, or in vitro tissue engineered decellularized matrices (Hoshiba et al., 2010). Decellularized scaffolds from various intact tissues, such as decellularized heart valves (Fang et al., 2007; Schmidt et al., 2007), blood vessels (Shimizu et al., 2007; Yazdani et al., 2009), small intestinal submucosa (SIS) (Badylak et al., 1989; Badylak et al., 1995; Kropp et al., 1995; White et al., 2005), lung, trachea, skin, nerve, cornea and other tissues and organs (Hoshiba et al., 2010), have all been investigated.

Some of the studies performed on decellularized matrices have provided evidence that complete removal of cells with preservation of not only the ECM microstructure but also its biologically active components, such as the growth factors,

22 is achievable (Badylak, 2007; Badylak et al., 2009). Importantly, tissue engineered decellularized scaffolds also did not elicit an immune response when used in vivo (Bloch et al., 2011).

One of the main challenges that may impair the clinical utilization of decellularized tissues and organs is the deterioration of the mechanical properties after the decellularization processes. This may decrease the regenerative potential of the decellularized ECM, especially if high mechanical properties are required as is the case with bone, cartilage and tendon applications (Elder et al., 2009; Haag et al., 2012; Murase et al., 2006; Tu et al., 2007; Yin et al., 2013). However, it has been reported that following decellularization, certain treatments could enhance and/or restore mechanical properties, mainly by crosslinking approaches (Grandi et al., 2011; Haag et al., 2012).

One of the novel methods of fabricating biomimetic scaffolds was introduced by Takeda and Xu (2014), whereby ECM of heart and liver tissues was reconstituted using a combination of homogenization and enzymatic digestion. These ECM solutions were reconstituted into 2D films by using crosslinking reagents. Also, in the same study a mechanically robust 3D biomaterial scaffold was fabricated by freeze- drying of the same ECM solution. It was proposed that these 2D and 3D bio-scaffolds have considerable potential to be utilized in tissue engineering and regenerative medicine applications (Takeda and Xu, 2014).

A recent study utilized an autologous bone marrow mesenchymal stem cell- derived ECM for cartilage tissue engineering. A 3D porous scaffold was fabricated via cross linking and freeze-drying techniques. Articular chondrocytes were seeded into the porous scaffold to form new cartilage after 4 weeks of in vitro culture and 3 weeks of in vivo implantation, which was claimed to be thick and homogenous in structure (Tang et al., 2013).

These type of approaches to fabricating decellularized scaffolds based on cell derived ECM are gaining more attention in recent years. In line with this direction of research, tissue engineered decellularized matrix was achieved by seeding mesenchymal stem cells onto polyester urethane (PEU) foams, then decellularizing the construct using a freeze thawing cycle followed by perfusion with phosphate

buffered saline (PBS). The authors reported DNA removal of up to 96%, while preserving the organic and inorganic components of the ECM (Sadr et al., 2012).

Utilizing a similar strategy, Syedain et al. developed a tissue engineered cell derived 3D leaflet construct for heart valve replacement using contractile cells seeded on a fibrin based scaffold to produce mature ECM and then the cells were removed using 1% SDS and 1% Triton X-100. they showed that ECM integrity was maintained after decellularization, and showed good potential for recellularization in vitro, after seeding hMSC on top of the decellularized leaflets (Syedain et al., 2013).

In a similar study, the concept of having a tissue engineered heart valve as an off-the-shelf product was introduced based on biodegradable synthetic polymer (non- woven polyglycolic-acid meshes) scaffold that was seeded with ovine vascular derived cells and left in culture for four weeks. The scaffolds were then decellularized with a mixture of chemicals and enzymes. The authors reported that decellularization did not negatively affect the collagen structure of the ECM, and the decellularized constructs supported recellularization when reseeded with mesenchymal stem cells (Dijkman et al., 2012).

Notably, cell derived decellularized ECM constructs have also been used in cancer research as models for investigating metastatic cell behavior. For example, approach has been used to increase the understanding of breast cancer cell interactions within an engineered bone micro-environment (Taubenberger et al., 2013; Thibaudeau et al., 2014a; Thibaudeau et al., 2014b) and also to investigate the mechanism of interactions between human prostate cancer cells and human bone (Hesami et al., 2014) by creating bone like mineralized tissue derived from human osteoblasts.

The influence of decellularized matrix on gene expression and cell differentiation has not been widely reported in the literature. The few studies that have assessed gene expression and cell differentiation have mainly utilized chondrogenic and osteogenic cells seeded into human mesenchymal stem cell (hMSC) derived ECM. It has been shown that cell derived ECM is capable of inducing seeded progenitor cells to differentiate and express genes native to the original tissue, suggesting that hMSC – derived ECM scaffolds are promising candidates for cartilage and bone tissue engineering (Choi et al., 2014; Thakkar et al., 2013).

The utilization of tissue engineered cell derived ECM using fibrous cell sheet technology has been only reported very recently in two studies, one of which was published out of the present PhD study (Farag et al., 2014; Xing et al., 2015). The approach was based on developing 2D decellularized fibrous cell sheets using optimized decellularization techniques that preserve the structural and functional components of cell sheets and to have the potential of recellularization in vitro.

The development of tissue engineered decellularized constructs for periodontal regeneration applications using cell sheet technology is the main focus of this PhD.

Research hypothesis:

The hypotheses of this thesis are:

1) Periodontal ligament cell sheets can be decellularized to fabricate a cell derived extracellular matrix that preserves its structural and biological properties after the decellularization process, and has the potential for recellularization both in vitro and in vivo.

2) The decellularized periodontal ligament cell sheet constructs have the capability to induce cell differentiation and trigger the expression of periodontal and bone tissue markers in vitro.

3) The decellularized constructs are biologically compatible and have the potential to facilitate periodontal regeneration in vivo.

Aims:

1. Produce and optimize a tissue engineered cell derived extracellular matrix using periodontal ligament cell sheets that retains structural integrity and biofunctionality, and can support allogenic cell recellularization.

2. Characterize and compare the efficacy of different decellularization protocols in terms of their ability to preserve the biological and structural properties of human periodontal ligament (h-PDL) cell sheet constructs.

3. Assess the effect of the decellularized cell sheet constructs on allogenic MSC and PDL cell differentiation and gene expression in vitro.

4. Assess the in vivo biocompatibility and periodontal regenerative potential of tissue engineered decellularized PDL cell sheet constructs.

Potential Significance and Innovation

The goal of any regenerative procedure is to restore the original structure and function of damaged tissue. The objective of the research reported in this thesis is to use periodontal ligament cell sheets to produce a cell-derived extracellular matrix which is combined with a polycaprolactone scaffold to form decellularized periodontal ligament cell sheet constructs. I believe that this technology will have the potential to be commercialized as an ‘off-the-shelf’ product for clinical applications.

Chapter 2

Decellularized periodontal ligament cell sheets with recellularization potential

Abstract

The periodontal ligament is the key tissue in facilitating periodontal regeneration. This study aimed to fabricate decellularized human periodontal ligament cell sheets for subsequent periodontal tissue engineering applications. The decellularization protocol involved the transfer of intact human periodontal ligament cell sheets onto melt electrospun polycaprolactone membranes and subsequent bi-directional perfusion with NH4OH / Triton X-100 and DNase solutions. The protocol was shown to remove 92% of DNA content. The structural integrity of the decellularized cell sheets was confirmed by a collagen quantification assay, immunostaining of human collagen type I and fibronectin and scanning electron microscopy. ELISA was used to demonstrate the presence of residual basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF) in the decellularized cell sheet constructs. The decellularized cell sheets were shown to have the ability to support recellularization by allogenic human periodontal ligament cells. This study describes the fabrication of decellularized periodontal ligament cell sheets which retain an intact extracellular matrix and resident growth factors, and can support repopulation by allogenic cells. The decellularized hPDL cell sheet concept has the potential to be utilized in future ‘off-the-shelf’ periodontal tissue engineering strategies. Key Words: tissue engineering, periodontal ligament, cell sheet, decellularization, recellularization, polycaprolactone.

Introduction

Periodontal regeneration can be considered to be the ultimate goal of periodontal treatment. However, currently available clinical techniques are unpredictable (Esposito et al., 2009; Needleman et al., 2006). Given that the periodontium is a complex structure involving soft and hard tissues, a tissue engineering approach is inherently suited to periodontal regeneration as it has the potential to facilitate the correct temporal and special arrangement of the multiple components required for periodontal regeneration (Bartold et al., 2000). The periodontal ligament has long been considered to be the key tissue required for periodontal regeneration (Nyman et al., 1982). Indeed, periodontal ligament cells have been shown to have superior regenerative properties compared to other cells derived from the periodontium, such as gingival connective tissue and alveolar bone cells (Dan et al., 2014; Tsumanuma et al., 2011). These findings suggest that periodontal ligament cells and the associated extracellular matrix retain important cues that facilitate periodontal regeneration. Cell sheet technology allowing the non-enzymatic harvesting of cultured cell with an intact extracellular matrix provides the opportunity to deliver periodontal ligament cells directly to the root surface. This approach has been successfully used to promote periodontal regeneration in a number of small and large animal models (Akizuki et al., 2005; Flores et al., 2008; Ishikawa et al., 2009; Vaquette et al., 2012). Whilst tissue engineered cell sheet technology is promising, there are underlying limitations in reaching clinical practice, such as the reliance on the patient’s own cells, as well as the need for cell culture facilities and associated technical expertise. The use of decellularized matrices has the potential to overcome these limitations. Multiple studies have shown that decellarized tissues and organs can retain their biological and mechanical properties with no immune response upon transplantation in vivo (Dijkman et al., 2012; Dong et al., 2009; Fang et al., 2007). Further, tissue engineered decellularized constructs prepared in vitro were shown to retain their structural integrity, maintain their molecular functionality, and enhance tissue regeneration following in vivo transplantation (Sadr et al., 2012). This study tested the hypothesis that decellularized human periodontal cell sheets can retain their extracellular matrix integrity and support recellularization with allogenic cell. The aim of the study was to utilize a decellularization protocol that was optimized for cell sheets, and assess the extracellular matrix structural integrity, growth factor retention and recellularization potential of the decellularized periodontal ligament cell sheets.

Materials and methods

Primary human periodontal ligament cells isolation and culture

Human periodontal ligament cells (hPDLC) were obtained using an established protocol (Ivanovski et al., 2001). Briefly, after obtaining institutional ethics approval (Griffith University Human Ethics Committee) and informed patient consent, extracted third molars were obtained from patients aged 17-30 years old. Diced periodontal tissues were obtained from the middle third of the roots, and explanted into 25 cm2 flasks (figure 2.1A). The cells were subsequently grown and propagated in 175 cm2 flasks with Dulbecco’s Modification of Eagle’s medium (DMEM) supplemented with 10% foetal calf serum (FCS), Penicillin (50units/ml) and Streptomycin (50µg/ml). Cells between the third and fifth passage were utilized.

Decellularization, immunostaining and confocal imagining of cultured cell monolayers

In order to characterize the structure of the cell monolayer extracellular matrix, the cells were seeded at a density of 2 x 104 cells on 13 mm diameter Thermanox coverslips (Thermo Scientific NuncTM, Australia) and grown for 9 days in culture medium supplemented with ascorbic acid (100 µg/ml). At the end of culture period, a cell monolayer was formed and decellularized according to a protocol originally developed by Beacham et al. (Beacham et al., 2007) and modified by our group. Briefly, 800 µl of NH4OH (20mM) and Triton X-100 (0.5%) were added to the cover slips for 20 minutes and kept at 37oC. The coverslips were then rinsed twice with 1 ml of sterile water and kept in phosphate buffer saline (PBS) at 4 oC. Fresh and decellularized cell monolayers were imaged using confocal laser microscopy. Antibodies against human Collagen I and Fibronectin (Life Technologies, Invitrogen) were used to visualize the extracellular matrix. 4',6-diamidino-2-phenylindole (DAPI, 5 µg/ml) and Phalloidin – tetramethylrhodamine B isothiocyanate conjugate (Phalloidin-TRITC, 0.8 U/ml, life technologies, Invitrogen) was utilized to stain the nuclei and the actin fibers.

The technique of immunostaining is as follows; the coverslips were fixed with a 4% paraformaldehyde solution at pH 7.4 (Sigma-Aldrich, Castle Hill, NSW, Australia) in phosphate-buffered saline (PBS) for 20 min and thereafter rinsed with PBS. The cells were then permeabilized for 5 min in Triton X-100 (0.2%) in PBS followed by 2 rinses in PBS. The samples were then incubated for 10 min in bovine serum albumin (BSA 1%, Sigma- Aldrich) in PBS. Solutions of primary IgG antibodies against human collagen I and fibronectin (Life Technologies, Invitrogen, Carlsbad, CA, USA) were diluted in BSA/PBS (1%); antibody solution was added to the designated coverslips and incubated for 45 min. Samples were rinsed with PBS and incubated for another 45 min protected from light in BSA (1%) in PBS containing fluorescently labelled secondary antibody Alexa 633 goat anti-mouse antibody (5 μg/mL, Alexa Fluor, Catalog #A-21126, Invitrogen), 4′,6-diamidino-2- phenylindole (DAPI, 5 μg/mL), and Phalloidin–tetramethylrhodamine B isothiocyanate conjugate (Phalloidin-TRITC, 0.8 U/mL; Life Technologies, Invitrogen). We obtained controls for non-specific staining by omitting the primary antibodies (figure 2.6). The coverslips were rinsed with PBS then mounted on a microscope slide with mounting medium (ProLong Gold, Invitrogen). Confocal imaging was done with a Leica SP5 microscope with excitation wavelength of 633 nm.

Melt Electrospun carrier membrane fabrication

Polycaprolactone (PCL, CAPA®6400 Perstorp UK) was utilized for fabricating the carrier membranes utilizing a house built melt electrospinner on a static flat collector (figure 2.1D, 2.1E & 2.1F). The PCL granules were loaded into a 2 ml syringe and melt electrospun at 6 kV, 95oC, at a feed rate of 20 µl/hr with a spinneret-collector distance of 4 cm. A biopsy punch was used to produce 5mm diameter membranes. To enhance the scaffold hydrophilicity, a 2M NaOH treatment was performed for 30 minutes followed by 5 rinses in ultrapure water. The membranes were sterilized by immersing them in ethanol for 30 min followed by a 30-min UV irradiation. The membranes were used to support the cell sheet during the handling and decellualrization process.

Cell sheet harvesting

HPDL cells were seeded in 24 well cell culture wells at a seeding density 5x104 cells/well in media supplemented with ascorbic acid (1000 µg/ml). For the first 48 hours, the ascorbic acid concentration was ten folds greater than the standard concentration in order to enhance early extracellular matrix formation. The cells were then grown for 19 days in media supplemented with ascorbic acid (100 µg/ml), which was changed every 48 hours. At the end of the 21 day culture period, the PCL scaffold was positioned in the center of the 24-well plate, and the borders of the cell sheet were gently detached from the base of the well and pulled toward the edges of the scaffold by means of sterile fine-curved tweezers. This secured the cell sheet onto the PCL membrane, and the scaffold-cell sheet was placed into a 24-well plate with the cell sheets facing upward. To allow for cell adhesion of the cell sheet onto the membrane, the construct was further incubated for 4 hr. To prevent the drying of the cell sheet, a 25-μL quantity of medium was placed on the cell sheet. At the end of the adhesion period, 1 mL of medium was added to the well, and the cell sheets on the PCL scaffold were further incubated for 24 hr.

Decellularization of cell sheets

A bi-directional perfusion system developed by our group (figure 2.1G & 2.1H) was utilized to decellularize the cell sheets. It was composed of a pump, 30 ml plastic syringe, 3 mm diameter silicone tube and 2 x 15 ml falcon tubes stacked on each other. Rapid prototyped polylactic acid (PLA) porous constructs (lay-down pattern 0/90/180, figure 2.1H) were used as separators to divide the falcon tubes into compartments and ensure the appropriate positioning and stability of the PCL scaffolds within the falcon tubes. Three scaffolds were placed in a sandwich pattern between two PLA constructs using sterile tweezers, with a maximum of 9 scaffolds decellularized at a time. The decellualrization

solution, consisting of 30mL of 20 mM NH4OH solution with 0.5% Triton X-100, was bidirectionally perfused though the scaffold for 60 min at a rate of 1000 ml/hr with a flow inversion every 50 seconds. This was followed by perfusion in a DNase I solution (100U/ml, o Invitrogen) at 37 C in CaCl2 (0.9mM) and MgCl2 (0.5mM) in sterile PBS for 60 minutes. The PCL membrane-cell sheet constructs were finally perfused with sterile water at 37oC for another 60 minutes.

Confocal imaging of cell sheets

The PCL membrane-cell sheet constructs were immunostained for human collagen I and human fibronectin before and after decellularization to assess the sheet integrity using the same technique described above.

DNA content

DNA content was measured for both fresh cell and decellularized samples using a Quant – iT PicoGreen kit (Life Technologies, Invitrogen®) after matrix digestion in proteinase K as described in a previous study (Vaquette et al., 2013). For each group, 6 biological replicates were utilized and each measurement was performed in triplicate. The detailed protocol is as follows; all samples were transferred to new sterile 24-well culture plates, sealed with parafilm, and placed at -80°C for 48 hr. The samples were thawed at room temperature, then transferred into Eppendorf tubes containing 300 μl of Proteinase K solution (Life Technologies, Invitrogen) diluted in PBE (1:50) for a final concentration of 0.5 mg/mL, then placed at 65oC for 8 hr in a thermo-mixer (Eppendorf® Thermomixer®, BioExpress, Hamburg, Germany). Samples were diluted in phosphate-ethylene diamine tetra-acetic acid buffer (PBE, 1:20). Standards were prepared with λ DNA standard (Quant–iTTM PicoGreen®, Life Technologies) by the dilution series method according to manufacturer’s instructions, then plated in black 96-well plates (MicroWell–NUNCTM, Thermo Fisher Scientific). PicoGreen dye solution prepared according to the manufacturer’s instructions (Quant–iT PicoGreen®), was added to each sample, protected from light, and incubated for 10 min. The fluorescence (excitation 485 nm, emission 520 nm) was measured by means of a fluorescence plate reader (BMG PolarStar®, Ottenberg, Germany). The standard curve of λ DNA was used to calculate the final DNA content of the samples.

Scanning electron microscopy (SEM) of cell sheets

Fresh and decellularized cell sheets were fixed in glutaraldehyde (3%) for 1 day, then gradually dehydrated in gradient concentrations of ethanol, after which osmium tetroxide (Sigma-Aldrich) was utilized as a post-fixation agent for 60 min. The samples were dried

overnight, subsequently mounted on adhesive stubs with cell sheets facing upward, and finally gold coated under vacuum for 3 min. SEM imaging was performed utilizing a FEI Quanta 200 microscope operating at 10 kV with a working distance range of 11 to 15 mm.

Growth factor ELISA assay

ELISA were used to detect basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF) levels in both fresh and decellularized samples. Growth factor extraction was performed by adding 300 µl of NaCl (2M) in 20 mM HEPES with EDTA protease inhibitor cocktail (Roche complete mini, Roche Applied Science, Indianapolis, IN) to each sample and incubated for 60 minutes at room temperature with gentle shaking on an orbital shaker. Samples were collected into 1.5 ml Eppendorf tubes and centrifuged at 2000 rpm for 5 minutes. Growth factor quantification was carried out using a Bioplex assay (Bio-Plex Pro, Bio-Rad) according to the manufacturer’s instructions.

Collagen quantification:

Collagen content in the cell sheets was measured in both fresh and decellularized samples, by means of a hydroxyproline assay kit (Chondrex, Inc., Redmond, WA, USA; catalogue #6017). The samples were moved to a sterile 1.5-mL vial, and 100 μl of distilled water was added. The samples were sonicated for 30 sec at 3-second pulse and 1-second rest, with 50% amplitude. A 100-μL quantity of 10 N HCL was added per sample and incubated for 24 hr at 120oC. The samples were left to cool and then were transferred to new micro- centrifuge tubes and spun at 10,000 rpm for 3 min. Standards were prepared with serial dilutions. Samples were used undiluted. A 10-μL quantity of 10X Chloramine T solution and 90-μL of solution A were added to each sample in a 96-well plate, where samples and standards were plated in duplicate, then incubated for 20 min. A 50-μL quantity of 2X DMAB solution plus 50 μl of solution B was added to each well and further incubated for 30 min at 60oC. Plates were read by means of a plate reader (BMG PolarStar®).

Recellularization of decellularized sheets

We assessed decellularized sheets for their recellularization potential by seeding allogenic hPDL cells on the top of the decellularized constructs at a seeding density of 5,000 cells/ scaffold. Recellularization was assessed over 3, 7, and 21 days with confocal imaging, SEM imaging, and DNA quantification by the PicoGreen assay, according to the same methods as described earlier. For comparison, the cells were also loaded onto a commercially available non-cross-linked collagen membrane (Bio-Gide®, Geistlich Pharma AG, Wolhusen, Switzerland).

Statistical analysis

Results were expressed as ± mean standard deviation and student T test was used to analyze the data. The significance level of the statistical analysis was set at p < 0.05.

device. (A) PDL tissue explanted and PDL cells growth. (B) HPDL cells forming cell sheet (CSC) seen by light microscopy. (C) Decellularized cell sheet seen by light microscopy. (D&E) Melt electro spinner. (F) PCL scaffold fabricated by random deposition. (G) Perfusion system. (H) Spacers formed of Polylactic acid (PLA) polymer constructs with PCL scaffolds with cell sheets in-between.

Results

Scanning electron microscopy

The incorporation of ascorbic acid in the media along with in vitro cell culture resulted in the deposition of a well-developed collagenous network and hence a mature cell sheet was formed. The sheets were thick enough after 3 weeks of culture to be mechanically harvested with fine curved tweezers. This permitted the harvesting and placement of the cell sheet onto a PCL melt electrospun scaffold (Figure 2.2A,B). Attachment of the cell sheet to the PCL scaffold was rapid, provided that the scaffolds were surface treated with sodium hydroxide to increase their hydrophilicity. It was found that a 24 hour period was sufficient for the cell sheet to become firmly adherent to the scaffold and withstand the subsequent fluid perfusion decellularization process. The SEM images revealed that both the fresh and the decellularized cell sheets remained intact and well attached to the PCL scaffold (Figure 2.2C and 2.2D). Higher magnification images of the decellularized samples demonstrated the presence of a fine network of collagen fibers with a morphology and structural integrity similar to that observed in the fresh cell sheet (Figure 2.2D I-III & 2.2D IV-VI).

Figure 2.2: Harvesting of fresh Human periodontal ligament (HPDL) cell sheet and scanning electron microscopy (SEM) showing fresh and decellularized PDL sheet. (A)

PCL scaffold of 5 mm diameter, after NH4OH treatment. (B) Fresh HPDL cell sheet attached to sterile PCL scaffold. (C) SEM image of decellularized PDL sheet on top of PCL scaffold. (D) Different SEM magnification of fresh (I-III) and decellularized (IV-VI) sheet.

Extracellular matrix characterization

Figure 2.3A and 2.3B displays representative immunostaining of hPDLC monolayers cultured on a cover slip, showing a well-developed network of fibronectin and collagen fibers. Upon decellularization, the components of the extracellular matrix formed by the monolayers were well preserved as shown in Figure 2.3C and 2.3D with no apparent alteration in their structural integrity when compared to the fresh matrices. Similarly, in the case of the mature cell sheets placed on the PCL membranes (Figure 2.3E and 2.3F), the decellularization protocol resulted in preservation of the quality and integrity of the extra-cellular matrix components (Figure 2.3G and 2.3H). Negligible traces

of DNA remnants (in blue) and actin filaments (in red) were detected in the decellularized sheets, indicating efficient removal of cellular contents using this decellularization protocol.

Figure 2.3: Immunostaining of human collagen type I and fibronectin. (A-D) Staining of cell monolayers on coverslips; (E-H) Staining of mature cell sheet – polycaprolactone constructs. Nuclei (DAPI) in blue, Actin filaments (Phalloidin) in red, human collagen type I and human Fibronectin in grey.

DNA quantification

DNA quantification confirmed the efficacy of the decellularization protocol in removing the cellular components, with 92% of DNA successfully eliminated from the hPDLC sheets as shown in Figure 2.4A.

Growth factors ELISA and collagen quantification

bFGF, VEGF and HGF were found to be retained in the decellularized sheets. As shown in Figure 2.4B, approximately 10% of the initial growth factor content in the fresh cell sheet remained after decellularization. Collagen quantification revealed increased collagen

content in the decellularized cell sheets, indicating that this decellularization method did not affect the amount of retained collagen (Figure 2.4C).

Figure 2.4: Comparison of DNA amounts, growth factor concentrations and collagen content of fresh and decellularized periodontal ligament cell sheet constructs. (A) DNA content before and after decellularization. (B) Growth factors retained in fresh and decellularized sheet, Fibroblast Growth Factor (FGF), Vascular Endothelial Growth Factor (VEGF), Hepatocyte Growth Factor (HGF). (C) Collagen quantification showing

preservation of collagen content after decellularization process. (***) = statistically significant where P<0.05.

Cell growth on decellularized sheets

The decellularized matrices were re-seeded with allogenic hPDL cells and cultured in vitro over 21 days. The recellularized constructs showed a gradual and significant increase in DNA content indicating that the matrices were capable of supporting cell adhesion and proliferation as shown in Figure 2.5C. The cells adopted a spindle-like morphology (black arrows) as seen by confocal laser microscopy (Figure 2.5A) and SEM (Figure 2.5B). Newly formed extracellular matrix was observed at the later time points (Figure 2.5A-III and 2.5-VI) indicating excellent cyto-compatibility of the decellularized substrate. We compared the decellularised sheet to a commercially available and well-established collagen membrane (Bio-Gide) and demonstrated that PDL cells proliferated better on the decellularized cell construct (Figure 2.5C & Figure 2.7).

Figure 2.5: Recellularization potential of the decellularized sheet after seeding with allogenic hPDL cells after 3, 7 and 21 days respectively. (A) Confocal imaging and immunostaining of human collagen type I (grey), Nuclei (blue), actin filaments (red). (B) SEM showing hPDL cells at different time points. (C) DNA quantification showing cell proliferation over 21 days.

Figure 2.6: Secondary antibody control for immunostaining. (A&B) Nuclei (DAPI) in blue, Actin filaments (Phalloidin) in red.

Figure 2.7: Seeding of hPDL cells on BioGide collagen membrane over 3, 7 and 21 days respectively. (A) BioGide membrane of 5 mm diameter. (B) SEM of membrane with no hPDL cells. (C) SEM showing hPDL cells on membrane at different time points. (D) Confocal imaging of hPDL cells showing nuclei (blue) and actin filaments (red).

Figure 2.8: SEM of multilayered (4 layers) hPDL sheets (A-C) hPDL multilayered cell sheets before and after decellularization.

Discussion

Decellularization is an emerging technology in regenerative medicine which is showing considerable potential in providing ‘off-the-shelf’ tissue engineered constructs for clinical use ((Weber et al., 2013). Most of the attention in this field has focussed on the decellularization of whole organs and tissues, with fewer reports on the decellularization of in vitro tissue engineered constructs. Notably, commercially available products based on decellularization approaches, such as Alloderm®, have been used in dental applications, namely mucogingival surgery (Cothren et al., 2004; Nevins et al., 2011; Salzberg et al., 2011; Sanz et al., 2009). The existing literature would suggest that decellularized organs, tissues

and constructs are viable options for tissue engineering applications, with considerable clinical potential (Tapias and Ott, 2014).

The harvesting of a mature cell sheet with an intact extracellular matrix is a key requirement for the decellularization protocol. Periodontal ligament cell sheets have been shown to support periodontal regeneration (Ishikawa et al., 2009), although a significant issue in the outcome is the ability to deliver and maintain an intact cell sheet on the tooth surface (Tsumanuma et al., 2011). To this end, we have previously utilised melt electrospun PCL membranes to successfully deliver and stabilize the cell sheets into experimental periodontal defects (Dan et al., 2014). Therefore in this study, prior to decellularization, the cell sheets were transferred to PCL membranes, which provided support during the decellularization process and allowed ease of handling of the decellularized construct. The PCL material was chosen as it is a clinical grade product, which would facilitate future translation of the decellularized construct to the clinic.

This study adopted a decellularization protocol that was originally used by Beacham et al. (Beacham et al., 2007). Changes to the protocol included the addition of DNase digestion as well as the bi-directional perfusion steps, which aimed to enhance the removal of cellular contents from the decellularized sheet, thus minimizing the possibility of eliciting an immune response upon future in vivo transplantation ((Brown et al., 2009). In the present study, we were successful to remove more than 92% of DNA content, which is comparable with the findings of Sadr et al. (2012), who showed 94% DNA removal following decellularization in a perfusion bioreactor and no adverse effects following in vivo implantation. Studies have demonstrated DNA removal above 99% with the use of sodium dodecyl sulfate (SDS) as the decellularization agent but this comes at the expense of extracellular matrix integrity (Syedain et al., 2013).

A key requirement of the decellularization process is that the integrity of the extracellular matrix is maintained so that it retains the biological cues that would lead to tissue specific cell differentiation of native cells upon in vivo transplantation. Type I collagen and fibronectin are two of the predominant proteins present in the native periodontal ligament tissue (Nanci and Bosshardt, 2006a), and we were able to demonstrate their presence and structural integrity following decellularization of both hPDL cell monolayers and the mature cell sheet – PCL membrane constructs. Further, collagen quantification showed increased 44

collagen content after decellularization. This suggests that most if not all of the original collagen was retained in the decellularized matrices, with the observation of increased concentration values consistent with other reports in the literature (Quint et al., 2011), and likely to be due to the removal of the cellular contents from the sheets. The results were consistent with other studies who have demonstrated intact extracellular matrix following the decellularization of tissue engineered constructs (Quint et al., 2011). Decellularized multilayered cell sheet (4 sheets) were investigated as shown in Figure 2.8. Despite the presence of what appear to be channels within this construct that may promote cell infiltration, the implantation of such a thick collagenous tissue network may be detrimental to initial cell infiltration and neovascularization in vivo compared to a single sheet which is more porous. Constructs utilized in this study, the handling and space maintenance properties of the PDL tissue are provided by the PCL scaffold, and hence a multi-layered sheet is not absolutely necessary.

It was also demonstrated that growth factors could be retained in the decellularized cell sheets, at levels which were approximately 10% of those detected in fresh cell sheets. These results were consistent with the findings of Reichert et al. (2010) who also identified growth factors in their decellularized matrices formed by primary human osteoblasts on a polymeric substrate. Notably, the growth factor concentrations obtained in our study were approximately 10 times higher than those reported by Reichert et al. (2010). This may be due to differences in methodology as Reichert et al. (2010) used a different cell type and culture method.

The allogenic hPDL cells were able to attach to the decellularized constructs as demonstrated by confocal imaging and SEM. Furthermore, the allogenic cells continued to proliferate and deposit extracellular matrix for the full course of the recellularization experiment, which ran for 21 days. This finding is consistent with another report where decellularized tissue engineered heart valves could be subsequently recellularized with allogenic mesenchymal stem cells (Syedain et al., 2013).

In conclusion, this study describes the fabrication of decellularized periodontal cell sheet constructs with extracellular matrix integrity and the ability to support allogenic cell repopulation. These constructs introduce the possibility of utilizing the instructive biological

signals that are constituent to periodontal ligament cells and the associated extracellular matrices as an ‘off-the-shelf’ tissue engineering application.

Chapter 3

Assessment of static and perfusion methods for decellularization of PCL membrane supported fibrous cell sheet constructs

Abstract

Decellularization aims to harness the regenerative properties of native extracellular matrix while removing immunogenic cellular material. The aim of this study was to evaluate different methods for efficient removal of cellular material while maintaining extracellular matrix structural integrity and growth factor retention. To this end, intact human periodontal ligament cell sheets were placed onto melt electrospun polycaprolactone (PCL) membrane that reinforced the cell sheets and facilitated the physical manipulation necessary during the various decellularization protocols. These cell sheet constructs (CSCs) were decellularized

under static and perfusion conditions using a) 20 mM ammonium hydroxide (NH4OH)/Triton X–100, 0.5% v/v; and b) sodium dodecyl sulfate (SDS, 0.2% v/v), both with and without the addition of DNase. Reagent free freeze – thaw cycling (F/T) with perfusion was also assessed. The structural integrity of the decellularized CSCs was assessed using a collagen quantification assay, immunostaining of human collagen type I and scanning electron microscopy. Residual basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF) and hepatocyte growth factor (HGF) were assessed with Bio-plex assays. DNA removal in the absence of DNase was higher under static compared to perfusion

conditions (NH4OH/Triton X–100 – 53% ± 7.5 & 0% ± 7.8, SDS – 75% ± 3.7 & 40% ± 3.8). However, after DNase treatment, there were no differences between static and perfusion decellularization with both resulting in virtually 100% DNA removal. Using the F/T technique, DNA elimination was 30.5% ± 11, and 90.6% ± 1.6 after DNase treatment. Collagen content was not affected by the various decellularization techniques, except after SDS treatment without DNase (34.5% ± 3.1 & 14.4% ± 2 for static and perfusion

respectively). Structural integrity was preserved after NH4OH/Triton X–100 and F/T treatment, while SDS altered the extracellular matrix structure. Growth factor amounts were reduced after decellularization with all methods, with the greatest reduction to virtually

undetectable amounts following SDS treatment, while NH4OH/Triton X-100 and DNase treatment resulted in approximately 10% retention. This study showed that treatment with

NH4OH/Triton X-100 and DNase solution was the most efficient method for DNA removal and the preservation of extracellular matrix integrity and growth factors retention.

Introduction

Tissue engineered cell sheet technology has gained attention as a promising technique in the field of regenerative medicine (Dan et al., 2014; Flores et al., 2008a; Flores et al., 2008b; Ishikawa et al., 2009; Vaquette et al., 2012; Zhao et al., 2013; Zhou et al., 2007). However there are a few underlying limitations hindering this technology from being applicable in clinical practice. A significant issue is the reliance on an appropriate cell source in terms of functionality and adequate cell numbers, with autogenous sources hampered by patient morbidity and heterogeneity in regenerative capacity, while allogeneic sources are associated with safety concerns. There are also the issues of dedicated cell culture facilities, technical expertise, transport and associated costs.

Decellularization is a strategy that could be utilized to overcome the potential limitations to applying cell sheet technology to the clinical setting by removing the necessity of implanting constructs containing viable cells. The effectiveness of decellularized tissues and organs has been widely reported in regenerative medicine applications, showing that biological and mechanical properties are retained following the decellularization process without eliciting an adverse immunogenic response when implanted in vivo (Burk et al., 2014; Nonaka et al., 2014; Syed et al., 2014; Weber et al., 2013; Xiong et al., 2013; Zhang et al., 2013). Aside from the use of native decellularized tissues and organs, tissue engineered decellularized constructs prepared in vitro have been shown to retain their structural integrity and maintain their molecular functionality (Elder et al., 2009), as well as enhance tissue regeneration when used in vivo (Sadr et al., 2012).

Decellularization has the potential to have significant implications for the commercialization of tissue engineered constructs by facilitating the development of ‘off-the- shelf’ products. Indeed, decellularized allografts and xenografts such as Alloderm® and Mucograft® are already commercially available for clinical use in a variety of fields, including cosmetic surgery, maxillofacial surgery and periodontics (Bloch et al., 2011; Dijkman et al., 2012; Fang et al., 2007; Flynn et al., 2006; Hoshiba et al., 2010; Shimizu et al., 2007; White et al., 2005; Yazdani et al., 2009).

The use of fibrous cell sheets may be particularly suited to the regeneration of soft- hard tissue interfaces, such as tendons and ligaments in the field of orthopaedics and periodontics (A et al., 2015; Ahn et al., 2015; Akizuki et al., 2005; Dan et al., 2014; Ma et al.,

2012; Vaquette et al., 2012; Vaquette et al., 2013). However, the utilization of fibrous decellularized cell sheets has only recently been reported in the literature (Farag et al., 2014; Xing et al., 2015) and presents some unique challenges, with the inherent fragility of the sheets presenting issues with handling and delivery. These challenges are accentuated in the context of decellularization, which is a technique which requires considerable handling and manipulation. The use of a thin electrospun membrane produced from a material such as PCL has been shown to support fibrous cell sheets used in the regeneration of soft-hard tissue interfaces such as that between periodontal ligament (PDL) and dentine (Costa et al., 2014; Dan et al., 2014).

Various approaches have been described for tissue decellularization, including chemical, physical and enzymatic treatments (Badylak et al., 2009). For a decellularization protocol to be efficient, a combination of the aforementioned approaches is usually required (Gilbert et al., 2006; Syed et al., 2014). Sodium dodecyl sulfate (SDS), an ionic detergent, and Triton X-100 (t-octylphenoxypolyethoxyethanol), a non-ionic detergent are widely used in many decellularization protocols for their cell lysis capacity (Weymann et al., 2015; Wu et al., 2015). Each of these approaches have advantages and disadvantages, with specific decellularization approaches chosen on the basis of their suitability for a particular tissue engineering application. Given that fibrous cell sheets present specific challenges, it is prudent that the most suitable approach is selected for decellularization of cell sheet constructs.

To this end, this study aimed to investigate different methods for the decellularization of PCL membrane supported PDL fibrous cell sheets under both stationary and dynamic fluid conditions, in order to identify the most efficient technique for the removal of cellular contents, which at the same time maximizes extracellular matrix integrity and growth factor retention.

Materials and methods

Membrane fabrication via melt electrospinning writing

Customized membranes were fabricated using medical grade polycaprolactone (mPCL, Purasorb PC 12, Corbion-Purac) via melt electrospinning direct writing (Brown et al., 2011). The polymer was melt electrospun at a temperature of 100oC, a feed rate of 20 µL/hr, a voltage of 10kV and a spinneret collector distance of 2 cm. The translational speed of the collector was set at 250 mm/min in order to obtain straight fibers and a square wave pattern was utilized for fabricating a scaffold composed of alternating series of layers oriented at 90°. The membranes were sectioned into 5mm discs. In order to increase their hydrophilicity, the melt electrospun membranes were etched with 2M NaOH for 30 minutes o at 37 C followed by 5 rinses in ultrapure water. The membranes were sterilized by exposure to 70% ethanol for 30 minutes followed by evaporation under the cell culture hood with another 30 minutes of UV irradiation. These discs were utilized as a support membrane in order to harvest the cell sheet and facilitate their handling.

Primary human periodontal ligament cells (h-pdl cells) isolation and proliferation

Human periodontal ligament cells (hPDLC) were obtained according to an established protocol, as previously described (Farag et al., 2014; Ivanovski et al., 2001). Briefly, explants were obtained from diced periodontal ligament tissue sourced from the middle 1/3 of extracted healthy teeth. The cells were grown to confluence and passaged using 0.05%

Trypsin and expanded into 175 cm2 flasks. Cells between the 3rd and 4th passages were used in this study.

Cell sheet harvesting

For the preparation of the cell sheets, the h-PDLCs were seeded in 24 well plates with a seeding density of 5x104 cells/well. For the first 48 hr, the ascorbic acid (catalogue number: A4403 - L-Ascorbic acid, Sigma-Aldrich) concentration was 1000μg/mL to enhance extracellular matrix formation (Beacham et al., 2007). The cells were then grown for 19 days in media supplemented with a lower ascorbic acid concentration (100μg/mL); the media was

changed every 48 hrs. At the end of the 21 days of culture the cells had deposited sufficient ECM in order to enable the handling of the cell sheet. In order to harvest the cell sheet, a PCL melt electrospun membrane was placed in the center of the well and the borders of the cell sheet were gently detached from the base of the well and folded over the edges of the membrane using sterile tweezers. The resultants cell sheet constructs (CSCs) were placed in expansion media for 24 hours with the cell sheets facing upward for allowing cell sheet adhesion onto the scaffold.

Decellularization protocols

Various decellularization methods were utilized and these techniques involved either flow perfusion, or static conditions with and without the utilization of DNase.

1. Static decellularization: The CSCs were decellularized by a static method whereby chemicals were added directly onto the construct in a 24 well plate: a) Ammonium Hydroxide (NH4OH)/Triton X-100

The CSCs were rinsed once with PBS and incubated into 20mM NH4OH (320145 SIGMA-ALDRICH) with 0.5% v/v Triton X-100 (1 mL solution per scaffold, 93443 SIGMA) for 30 minutes at 37oC. Thereafter, the decellularizing solutions were replenished and another 30 min incubation was performed prior to rinsing the scaffold three times using PBS at 37 oC. This method was originally adopted from the technique used by Beacham et al. (Beacham et al., 2007).

b) Ammonium Hydroxide (NH4OH)/Triton X-100 + DNase

The CSCs were decellularized using 20mM NH4OH with 0.5% v/v Triton X-100 (1 mL solution per scaffold) for a total of 60 minutes at 37oC as described above. This was followed by immersion in 1 mL DNase I solution (100U/ml, Catalog number: 18047-019 Invitrogen) in CaCl2 (0.9 mM) and MgCl2 (0.5 mM) in sterile PBS at 37oC for another 60 minutes before a final rinsing step using PBS at 37 oC.

c) Sodium dodecyl sulfate (SDS, 0.2% v/v, catalogue number: 05030 SIGMA)

The CSCs were rinsed with PBS and incubated into SDS (0.2% v/v in 1 mL solution per scaffold) for 30 minutes at 37oC. Thereafter, the decellularizing solutions were replenished and another 30 min incubation was performed prior to rinsing the scaffold using PBS at 37 oC.

d) Sodium dodecyl sulfate (SDS, 0.2% v/v +DNase)

The scaffolds were decellularized using SDS (0.2% v/v in 1 ml solution per scaffold) for a total of 60 minutes at 37oC, as described above. This was followed by a DNase step whereby the CSCs where immersed in 1 ml DNase solution (100U/ml) for another 60 minutes and incubated at 37oC before a final rinsing step using PBS at 37 oC.

All samples were then rinsed gently three times using ultrapure sterile water, then kept overnight in PBS at 4°C.

2. Perfusion Decellularization

A perfusion system bioreactor was designed in house (Figure 3.1A & 3.1B), and consisted of a series of spacers to separate the constructs, as well as silicon tubes, 0.2 µm filters and an infusion/withdrawal syringe pump. The chambers and its components were designed with CAD software and additive manufactured using an inkjet 3D printer (Objet30 Pro Desktop, Stratasys) with 16 microns deposited layers of an acrylic resin (Verowhite Plus 835, Stratasys). Dipping the chambers in acetone resulted in a water tight enclosure. An infusion/withdrawal syringe pump was used to perfuse the decellularization solutions through the scaffolds, hence allowing for a homogeneous decellularization. Similarly to the static decellularization, chemicals or detergents were utilized using the perfusion system. An additional methodology involving freezing and thawing was also assessed under perfusion. Table 3.1. summarizes the different methods of decellularization of cell sheets performed in this study.

Table 3.1. Decellularization methods used in this study

Decellularization method Static / Perfusion NH4OH + Triton-X 100 Both NH4OH + Triton-X 100 + DNase Both

SDS Both SDS + DNase Both Freeze/ Thaw Perfusion Freeze/ Thaw + DNase Perfusion

Figure 3.1. Decellularization perfusion system. A, The perfusion system composed of the bi-directional syringe pump connected to the perfusion chambers through silicon tubes. B, decellularization chamber components comprising of the inner carrier and the housing chambers. The chamber was CAD designed and 3D printed using photocurable material.

a) Perfusion + NH4OH /Triton X-100

The CSCs were rinsed once with warm PBS at 37oC, and then placed in the decellularization chambers with a maximum of 11 constructs per chamber. The CSCs were perfused in 30 mL of 20 mM NH4OH solution with 0.5% v/v Triton X-100. Bi-directional perfusion of the constructs was performed for 60 min at a rate of 1,000 mL/hr, with a flow inversion every 50s, followed by a final bidirectional perfusion in 30 mL sterile water at 37oC for another 60 min.

b) Perfusion + NH4OH/Triton X-100 + DNase

The decellularisation was performed as described above with an added DNase step involving perfusion in 30mL of DNase I solution (100 U/mL, Invitrogen) in CaCl2 (0.9 mM) o and MgCl2 (0.5 mM) in sterile PBS at 37 C for 60 min and finally perfused with sterile water at 37oC for another 60 min (Farag et al., 2014).

c) Perfusion +Sodium dodecyl sulfate (SDS, 0.2% v/v)

The same method as described in the previous section (a) with a similar bi-directional perfusion pattern.

d) Perfusion + Sodium dodecyl sulfate (SDS, 0.2% v/v) + DNase

Here again, a DNase step was performed (described in section b) prior to a final rinsing step with sterile water at 37oC for another 60 min under perfusion.

e) Thermal freezing/thawing cycles (F/T)

This method was originally developed by Sadr et al. (Sadr et al., 2012), and was implemented in the present study with minor modifications. The CSCs were initially rinsed with warm PBS. They were placed into 1.5 mL cryotubes (Thermo Scientific™) and three successive cycles of freezing/thawing (F/T) were subsequently performed. The freezing/thawing step consisted in immersing the cryotubes into liquid nitrogen for 5 minutes, and then transferring them to a 37oC water bath for 5 minutes. After each step of F/T, the CSCs were rinsed for one minute using warm ultrapure water to hypotonically lyse remaining cells. Finally, scaffolds were bidirectionally perfused for 60 minutes in PBS at 37oC.

e) Thermal freezing/thawing cycles (F/T) + DNase

Perfusion with DNase was added as an extra step prior to final rinsing via PBS perfusion, as a modification to the original Sadr et al. technique outlined.

The CSCs from all experimental methods were kept in PBS overnight at 4oC prior to subsequent evaluation.

Confocal imaging of cell sheets

The CSCs were immunostained using monoclonal antibodies against human Collagen I (Catalogue #: 63170 MP Biomedicals) to visualize, assess and compare the extracellular matrix and cell sheet integrity of the different decellularization protocols. 4', 6-diamidino-2- phenylindole (DAPI, 5 μg/ml) and Phalloidin – tetramethylrhodamine B isothiocyanate conjugate (Phalloidin-TRITC, 0.8 U/ml, life technologies, Invitrogen) were utilized to stain the nuclei and the actin fibers respectively. Briefly, samples were fixed with a 4% paraformaldehyde solution at pH 7.4 (Sigma-Aldrich, Australia) in phosphate buffer saline (PBS) for 20 minutes and thereafter rinsed with PBS. The cells were then permeabilised for 5 minutes in Triton X-100 (0.2%) in PBS followed by two rinses in PBS. The samples were then incubated for 10 minutes in 1% bovine serum albumin (BSA, Sigma-Aldrich, Australia) in PBS. Primary monoclonal isotype mouse IgG antibody against human Collagen I was diluted in BSA/PBS (1 %) with a dilution ratio of 1:200. The antibody solution was added onto the samples and incubated for 45 minutes at RT. The samples were rinsed with PBS and incubated in BSA (1%) in PBS containing the secondary Alexa 633 goat anti-mouse antibody (5μg/ml, Alexa Fluor, A-21126 Invitrogen), DAPI at 5 μg/ml) and Phalloidin-TRITC at 0.8 U/ml in dark for another 45 minutes. Controls for non-specific staining were obtained by omitting the primary antibody. Confocal imaging was undertaken with a Nikon Eclipse microscope (Nikon Eclipse 50iPOL, QUT Central Analytical Research Facility).

Scanning electron microscopy of cell sheets

Fresh and decellularized CSCs were fixed in 3% Glutaraldehyde for 1 day, and then samples were gradually dehydrated in Ethanol concentrations of 60-100% for 20 minutes in each step, then a post-fixation step with osmium tetroxide (Sigma-Aldrich, Australia) was

performed for 60 minutes. The samples were left to dry overnight, then mounted on adhesive stubs with cell sheets facing upwards, gold coated under vacuum for 3 minutes then Imaged using FEI Quanta 200 SEM.

Quantification of DNA

DNA content was measured in both fresh and decellularized CSCs. The specimens were transferred to a new sterile 24 well plate, sealed with parafilm and placed at -80oC for 48 hours. The samples were thawed at room temperature then transferred to Eppendorf tubes containing 300 µL of Proteinase K solution (life technologies- Invitrogen) diluted in PBE 1:50 for a final concentration of 0.5 mg/mL, then placed at 65oC for 8 hours in a thermomixer (Eppendorf thermomixer-BioExpress). Samples were diluted 1:20 in PBE; a standard curve was prepared with λ DNA standard (Quant – iT PicoGreen) using the dilution series method according to the manufacturer’s instructions. Test and standard solutions (100µL) were plated in triplicate in a black 96 well plate (MicroWell –NUNC, Thermo Fisher Scientific). Thereafter, 100uL of PicoGreen dye solution prepared according to the manufacturer’s instructions was added to each well and incubated for 10 minutes in the dark.

A fluorescence plate reader (BMG PolarStar, Ottenberg, Germany) with excitation 480nm and emission wavelengths of 480nm and 520 nm respectively was used to measure the fluorescence. A standard curve with known DNA concentrations was used to calculate the final DNA content in each sample.

Collagen quantification:

Collagen content in the CSCs was measured in fresh and decellularized samples using a Hydroxyproline assay kit (Chondrex, Inc. - catalog#6017) according to the manufacturer’s instructions. Briefly, the samples were placed in sterile 1.5 mL Eppendorf tubes and 100 μL of distilled water was added. The samples were sonicated for 30 seconds in cycles of 3 seconds pulse and 1 second rest with an amplitude of 50 %. 100 μl of 10N HCL was then added and incubated for 24 hours at 120°C. The samples were cooled down, transferred to micro centrifuge tubes and subsequently centrifuged at 10,000 rpm for 3 minutes. A standard curve was prepared from known collagen concentration solutions.

The samples were used undiluted. 10 μL of 10X Chloramine T solution and 90 μL of Chloramine T dilution buffer were added for each sample in a 96 well plate, where samples and standards were plated in duplicates, then incubated for 20 minutes. 50 μL of 2X Dimethylaminobenzaldehyde (DMAB) solution added to 50 μL of DMAB dilution buffer was then added to each well, incubated for 30 minutes at 60°C, and the optic density was evaluated using a plate reader (BMG PolarStar®) at a wavelength of 530-560 nm. Hydroxyproline levels (µg/mL) were determined using a standard curve prepared from reagents included in the assay kit, according to the manufacturer’s instructions.

Growth factor Bioplex assay

Basic fibroblast growth factor (FGF), vascular endothelial growth factor (VEGF) and Hepatocyte growth factor (HGF) levels in fresh and decellularized CSC samples were quantified using a multiplexed immunoassay (Bio-Plex ProTM-Bio-Rad). Growth factor extraction was carried out as previously described (Farag et al., 2014; Reichert et al., 2010; Thibaudeau et al., 2014b) whereby 300 µL of 2M NaCl in 20mM HEPES with EDTA protease inhibitor cocktail (Roche complete mini, Roche Applied Science, Indianapolis, IN) was added to each sample then incubated for 60 minutes at room temperature on an orbital shaker. Solutions from each sample were collected in 1.5 mL Eppendorf tube and centrifuged at 2000 rpm for 5 minutes then kept in -80°C till the time of the analysis. Assay for growth factors was done using a Bio-plex system (Bio-Plex ProTM-Bio-Rad) according to instruction manual provided by the manufacturer. Briefly, samples were placed on ice and gradually thawed to room temperature. Bio-Plex Pro™ Reagent Kit with Flat Bottom Plate (#171304070M) was used for the analysis. 50 µL of magnetic beads were added to selected wells of a 96 well plate provided by the manufacturer and washed twice with 100 µl wash buffer using a handheld Magnetic Washer (Bio-Plex® #171020100). Quantitative analysis was carried out using standards prepared in a series of dilutions (group I human cytokines (#171-D50001) for basic fibroblast growth factor (bFGF) and vascular endothelial factor (VEGF) and group II human cytokines (#171-D60001) for hepatocyte growth factor (HGF)). Briefly, 50 µL of standards, blank, and samples (undiluted) were plated in duplicate and incubated at room temperature with shaking at 850 rpm for 60 minutes. The plate was then washed three times with 100 µL wash buffer, and 25 µL of detection antibody was added to the wells and incubated for 30 minutes at room temperature protected from light on a shaker

at 850 rpm. After being washed again with wash buffer three times, 50 µL of Streptavidin- PE was added to each well, left for 10 minutes at room temperature protected from light on a shaker, followed by a final washing step. Finally, 125 µL of assay buffer was added for 30 seconds and the plate was read using a Bio-Plex suspension array system (Bio-Plex® 200 System) and concentrations were obtained in picograms per mL.

Statistical analysis

Results were expressed as mean ± standard error of mean from at least five biological replicates and one-way ANOVA test was used to analyze the data. The significance level for the statistical analysis was set at p < 0.05.

Results

DNA removal and DNase effect on decellularization

Static decellularization was found to be generally more efficient in the elimination of DNA contents of the CSCs when compared to the perfusion technique (Figure 3.2a). Treatment with SDS resulted in 75% and 40% of DNA removal using static and perfusion

decellularization respectively. The NH4OH/Triton X-100 protocol eliminated only 53% of DNA using the static method, while DNA elimination was almost negligible with the use of perfusion. DNA removal was significantly improved when DNase was added as an extra step after the initial decellularization. DNA was not detected in both the static and perfusion techniques utilizing the SDS-DNase combination, thus indicating complete DNA removal.

NH4OH/Triton X-100 combined with DNase eliminated all DNA content using static decellularization and up to 97% with the perfusion method. The freezing thawing (F/T) technique did not achieve adequate DNA elimination (approx. 31% of DNA content was removed) but with the addition of a DNase step, DNA elimination was significantly improved to around 91%.

Figure 3.2. DNA, collagen and retained growth factor quantification in decellularized human periodontal ligament cell sheets. A, Remnant DNA in the decellularized sheets was quantified by PicoGreen assay. B, Collagen content was measured using a hydroxyproline release assay. C-E, Residual growth factors detected in the fibrous cell sheets using a Bio-

plex assay. NH4OH/Triton X-100 + DNase using perfusion was superior in DNA removal, preservation of collagen and growth factor retention. Results having the same letter are not significantly different (P ≤ 0.05).

Structural preservation of decellularized sheets

Structural integrity of the cell-containing CSC extracellular matrix (Figure 3.3) and following the different decellularization protocols (Figure 3.4 & 3.5) was assessed using Scanning Electron Microscopy (SEM) and by immunostaining against human collagen type I that was visualized using confocal microscopy. CSCs treated with SDS with or without DNase were superior to all other techniques in terms of elimination of cellular and nuclear contents as seen by the lack of nuclear DAPI staining on the confocal imaging (figures 3.4I, 3.4M, 3.5I & 3.5M). However, SDS was noticeably destructive to the collagen fiber architecture when compared to both the NH4OH/Triton X-100 and Freezing–Thawing decellularization techniques. Disruption was manifested as alteration in the fibrous network structure of the cell sheet by fusion or coarsening of collagen fibers and/or loss of the fine fibrils normally seen in the untreated cell sheets (Figure 3.4L& 3.5L). Structural integrity was even more negatively affected in samples treated by the SDS–DNase combination, with more significant alteration in the fibrous structure of the extracellular matrix as demonstrated by

SEM (Figure 3.4P& 3.5P). Conversely, the NH4OH/Triton X-100 decellularization protocol preserved the extracellular matrix to a greater extent as demonstrated by SEM, but abundant cellular and nuclear debris was present (figure 3.4D& 3.5D). However, the addition of a DNase step retained the structural integrity of the CSC collagen fibers with significantly enhanced removal of cell debris as demonstrated at higher magnifications by SEM (Figure 3.4H& 3.5H). The Freezing/Thawing technique (F/T) appeared to preserve the collagen fibrous structure of the CSCs as seen in the immunostaining and SEM imaging, however significant cellular debris remained (Figure 3.5Q&3.5T), which was significantly reduced after the addition of a DNase step (Figure 3.5U&3.5X). F/T with or without the addition of DNase appeared to result in partial disruption of the CSC extracellular matrix (Figure 3.5Q & 3.5U). Generally, cellular and nuclear contents were greatly reduced or absent in samples treated with DNase, as seen by confocal microscopy (Figures 3.4E, 3.4M, 3.5E, 3.5M & 3.5U) and SEM (Figures 3.4H, 3.4P, 3.5H, 3.5P & 3.5X).

Figure 3.3. Confocal microscopy and SEM of fresh cell sheet. A, Collagen I immunostaining (White) of the fresh cell sheet, DAPI staining of nuclei (Blue) and phalloidin staining of the actin filaments (Red). B-D, SEM showing the fibrous network laid by the human periodontal ligament cells at different magnifications.

Figure 3.4. Confocal imaging and SEM of static decellularized cell sheets. A;E;I;M, Collagen I immunostaining (White), DAPI staining of nuclei (Blue) and phalloidin staining of the actin filaments (Red). B-D;F-H;J-L;N-P, SEM showing the fibrous network after static decellularization at different magnifications.

Figure 3.5: Confocal imaging and SEM of perfusion decellularized cell sheets. A;E;I;M;Q;U, Collagen I immunostaining (White), DAPI staining of nuclei (Blue) and phalloidin staining of the actin filaments (Red). B-D;F-H;J-L;N-P;R-T;V-X, SEM showing the fibrous network after perfusion decellularization at different magnifications.

Collagen content preservation

Generally, the various decellularization techniques did not seem to alter the collagen content in the CSCs with the exception of the static SDS treatment which reduced the collagen content by approximately 35%. NH4OH/Triton X100 and F/T cycles preserved most of the collagen content in the cell sheets. Also, DNase enzymatic treatment did not seem to have any significant effect on collagen preservation (Figure 3.2B).

Growth factors retention

Absolute quantities of all growth factors were significantly reduced in the decellularized CSCs irrespective of the method of decellularization. Static and perfusion methods showed comparable results in the percentage of growth factor (FGF, VEGF and HGF) retention in the decellularized sheets. Around 10% of the original growth factors (in a

fresh cell sheet) were retained when NH4OH/Triton X-100 was used under either static or perfusion conditions. Similar values were obtained with the F/T decellularization method. Retained quantities of FGF and HGF were either very low or undetectable in cell sheets decellularized using SDS. DNase treatment did not appear to affect the amounts of retained growth factors in the CSCs irrespective of the decellularization protocol with which it was combined (Figures 3.2C, 3.2D & 3.2E).

Discussion

Decellularization is a strategy that aims to harness the regenerative properties of native extracellular matrix while reducing potentially immunogenic cellular components. Cell sheet technology is a tissue engineering approach which has been recently used in the clinic to facilitate the in vivo delivery of cells together with an intact extracellular matrix (Hirabaru et al., 2015). Scaffold cell sheet constructs may be particularly useful for regenerating soft- hard tissue interfaces such as ligaments and tendons, and combined with appropriate decellularization methods have the potential to be developed as ”off-the-shelf” constructs for clinical use. In order to identify the best decellularization method that efficiently removes cellular material and maintains extracellular matrix integrity and growth factors retention, this study investigated the influence of various protocols on the structural integrity and growth factor retention in decellularized periodontal ligament fibrous cell sheets.

A well-recognized challenge of using cell sheets is the difficulty in handling, delivering and stabilizing an intact cell sheet (Flores et al., 2008b; Ishikawa et al., 2009). In order to improve cell sheet handling, we have previously utilized a thin membrane made of medical grade electro-spun polycaprolactone (PCL) as a carrier scaffold (Dan et al., 2014), which was also utilized in the current study. Given that the handling challenges are exacerbated during decellularization because of the additional manipulation that is required, a

support structure, such as the PCL membrane used in this study, should be considered an absolute requirement for cell sheet stabilization during the various decellularization protocols.

In this study, both the static and perfusion techniques were efficient in cellular and nuclear content removal, as measured by the amount of remaining DNA. The rationale for DNA removal is to avoid or at least minimise a possible immune response upon in vivo transplantation (Ivanovski et al., 2001). Although the static method was relatively superior in terms of DNA removal compared with the perfusion technique, neither technique achieved favourable DNA removal without the use of DNase I. Therefore, the use of DNase can be considered essential for effective DNA removal.

It was also noted that collagen integrity and content of the cell sheets was not different between the static and perfusion techniques. Further, there was no difference in growth factors retention between the static and perfusion methods when the same chemicals were used. Therefore, the efficiency of the decellularization and extracellular matrix structural and biological integrity appears to depend mainly on the nature of the chemicals, detergents and/or enzymes used, and to a lesser extent by the fluid mechanics during the decellularization process. However, it should be considered that the perfusion technique is more practical, technically less complicated and less time consuming when compared to the static technique, which would be important in the context of future upscaling for commercial applications. This is also the likely reason that the perfusion method appears to be favoured by recent studies utilizing decellularization of tissue engineered constructs (Bao et al., 2011; Syed et al., 2014).

The utilization of chemicals and detergents, with or without enzymes, plays a key role in the efficiency of decellularization. In this study, it was shown that more than 50% and 75% of DNA content were removed with SDS using static and perfusion decellularization protocols respectively, and almost complete DNA elimination (up to 99%) was achieved when enzymatic treatment using DNase was added. However, this came at the expense of growth factor retention. Indeed, growth factor concentrations were not only significantly decreased compared to other methods, but were also reduced to the extent that FGF could not be detected in the decellularized sheets. These results appear to contradict the findings of a recent study, which reported that dermal fibroblast cell sheets treated with SDS retained considerable amounts of FGF and VEGF (Xing et al., 2015). The differences might be

attributed to that study’s longer culture period and shorter period of decellularization, or the higher SDS concentration used in the present study. However, it must be noted that the decellularization conditions that resulted in the reported retention of growth factors following SDS treatment also resulted in the suboptimal removal of only between 45% (for low SDS 0.05 wt%) and 90% (for high SDS 0.5%) of DNA, as well as altered collagen structure as confirmed via immunostaining and SEM. This was consistent with a reported altered pattern of immunofluorescent staining for collagen type I and III in porcine bladders treated with SDS (Brown et al., 2005). These observations can be attributed to the denaturing effect of SDS on proteins of the extracellular matrix, and/or removal of other ECM components. It is also noteworthy that decellularized constructs treated with SDS were difficult to handle and decellularized sheets tended to detach from the PCL scaffolds most likely due to shrinkage artefacts. This is consistent with the findings of a recent study (Faulk et al., 2014) which evaluated the effect of different detergents on the basement membrane complex (BMC) and showed that SDS treatment could have detrimental effect on BMC integrity.

A thermal method for decellularization in the form of freeze-thaw (F/T) cycling was also evaluated. This approach has been favoured in several recent studies as it does not involve utilization of detergents and other potentially harmful chemicals, while achieving DNA removal of up to 96% with favourable preservation of the extracellular matrix integrity (Burk et al., 2014; Nonaka et al., 2014; Sadr et al., 2012). This method was found to be less efficient in decellularization of cell sheets in the current study unless an extra DNase treatment step was added. It is also noteworthy that the F/T method was more time consuming when compared to other techniques. Although collagen structure and content were preserved and growth factor retention was favourable, the gross overall integrity of the cell sheet was negatively affected by the F/T cycles as observed in both immunostaining and SEM imaging (Figure 3.5Q, R, U & V).

The combination of NH4OH/Triton X-100 and DNase was found to be the most efficient method of ECM structural preservation and DNA elimination irrespective of the whether the decellularization was carried out under static or perfusion conditions. Consistent with previous experiences with this method, DNA removal was efficient (up to 92%) without sacrificing the structural integrity of the ECM or the retention of the native growth factors in the decellularized sheets (Farag et al., 2014). This is consistent with the findings of a recent study which compared the efficacy of different protocols for preparation of extracellular matrix scaffolds derived from three-dimensional cell culture, and concluded that the

combined NH4OH and Triton X-100 treatment performed the best in terms of the removal of cellular components from the complexes (Lu et al., 2012).

In conclusion, this study assessed different protocols for the decellularization of

fibrous cell sheets, and showed that the combination of NH4OH and Triton X-100 together with a DNase treatment step was the most efficient method for DNA removal and preservation of extracellular matrix integrity and growth factors retention, irrespective of whether a static or perfusion approach was used. The perfusion protocol may have the added advantage for automatization and throughput upscalability that may facilitate future commercialization.

Chapter 4

The effect of decellularized periodontal ligament constructs on cell differentiation in vitro and periodontal regeneration in vivo.

Abstract

Cell derived extracellular matrix is well known to have the potential to induce differentiation of progenitor cells, thus guiding the fate of host cells to achieve regeneration. In this study, decellularized periodontal ligament cell sheets were assessed for their effect on human periodontal ligament and mesenchymal stem cell differentiation, as assessed by the gene expression of markers relevant to bone and periodontal ligament tissues as osteocalcin (OCN), osteopontin (OPN), alkaline phosphatase (ALPL), bone morphogenetic protein type 2 (BMP2), collagen type I and type III, tenascin C (TNC) and vascular endothelial growth factor (VEGF) at days 3, 7 and 14 post-seeding, and STRO-1 immunostaining. Further, the decellularized constructs were evaluated in a rat periodontal defect model to assess their biocompatibility and tissue integration, as compared to empty defects and fresh cell sheet constructs. Microcomputed topography (µCT) and histological assessment utilizing H&E, Mason Trichome and immunohistochemistry staining for OCN and OPN bone markers were performed to assess the regenerative potential of the different constructs at 2 and 4 weeks postoperatively.

There was up-regulation of bone marker gene expression in allogenic human periodontal ligament cells (hPDLC) seeded on top of the decellularized sheets especially on the 14th day post-seeding. However, mesenchymal stem cells did not express bone markers, but rather showed increased collagen marker expression at the earlier time points (days 3 and 7). STRO-1 expression by the MSCs decreased over the three time points when seeded on decellularized sheets, reaching the lowest expression on the 14th day. However in the absence of decellularized cell sheets, STRO-1 expression remained the same at all time points for cells seeded on polycaprolactone (PCL) scaffolds alone, suggesting no cell differentiation over the period of 14 days.

Histological assessment of the constructs inserted in the periodontal defects demonstrated the biocompatibility of the decellularized periodontal ligament sheets and showed their ability to integrate into the surrounding tissues with full recellularization and no signs of adverse tissue reactions. Taken together our data showed that decellularized periodontal ligament cell sheets are capable of inducing cell differentiation, upregulating the expression of bone markers in hPDL cells and have the potential to facilitate periodontal regeneration in vivo.

Introduction

Periodontal disease is a widespread condition that leads to the loss of a tooth’s supporting tissues, and may result in the eventual loss of the tooth (Armitage et al., 2003; Kawaguchi et al., 2005; Nyman et al., 1982).

Various attempts have been made to restore the lost periodontal structures using different reconstructive surgical methods such as guided tissue regeneration, which has been extensively covered in the literature (Cortellini and Tonetti, 2000; 2005), These techniques are sometimes combined with bone graft materials to enhance bone and periodontal tissue regeneration (Bashutski and Wang, 2009; Dori et al., 2008b; Lin et al., 2008).

However, currently available clinical techniques have not been able to fully restore the damaged periodontium mainly due to its complex structure (Britain et al., 2005).

Tissue engineering of the periodontium is a promising future treatment option for periodontal regeneration with numerous pre-clinical studies investigating this approach (Bai et al., 2011; Maeda et al., 2011; Reichert da Silva Assuncao et al., 2011). The introduction of cell sheet technology to the field of tissue engineering has been well established by many studies demonstrating its regenerative potential for bone, cartilage, skin and other tissues (Ahn et al., 2015; Gao et al., 2015).

Cell sheet technology has also been used in the field of periodontal regeneration and has been shown to support the regeneration of periodontal tissues such as cementum, periodontal ligament and alveolar bone in pre-clinical animal models (Akizuki et al., 2005; Bai et al., 2011; Dan et al., 2014; Vaquette et al., 2012).

Although the utilization of cell sheets in periodontal regeneration has provided promising outcomes, this technology has certain limitations that might hinder its future clinical application in the field of periodontics. For example the need for donor cells from each patient receiving this kind of treatment, the necessity of having cell culture facilities and possible additional cost due the need for expertise in this technology (Farag et al., 2014).

Tissue engineered cell derived extracellular matrices are gaining more attention in the field of regenerative medicine due to their ability to restore damaged or lost tissues such as

bone (Benders et al., 2013; Papadimitropoulos et al., 2015; Sadr et al., 2012), cartilage (Gawlitta et al., 2015) and tendons (Badylak, 2007; Badylak et al., 2009; Yin et al., 2013).

The utilization of decellularized constructs to induce regeneration has been assessed in a few studies, where it was shown to induce ectopic bone formation in animal models (Sadr et al., 2012).

In the current study, a novel tissue engineered construct using decellularized cell sheet technology was assessed for its biocompatibility and potential regenerative capacity in both in vitro and in vivo models. The effect of the decellularized sheet on cell differentiation and gene expression in both human periodontal ligament cells (hPDL) and human placental mesenchymal stem cells was assessed to determine the inductive potential of the decellularized matrices. Furthermore, the ability of the decellularized constructs to facilitate periodontal regeneration in vivo was assessed in a rat periodontal defect model.

Materials and Methods

Scaffold fabrication by melt electrospinning writing

Medical grade polycaprolactone (PCL) melt electrospun scaffolds were fabricated using melt electrospinning direct writing (Brown et al., 2011). PCL polymer was melt electrospun at feed rate of 20 µL/hr, at a temperature of 100oC with a voltage of 10kV and the spinneret collector distance was set at 2 cm. The collector was set at translational speed of 250 mm/min in order to obtain straight fibers. A square wave pattern was utilized for fabricating a scaffold composed of alternating series of layers oriented at 90° (figure 4.1A). Scaffolds to be used in the in vitro experiments were sectioned into 5mm discs using a 5 mm biopsy punch (Kai medical, Japan). For the in vivo experiment, scaffolds were cut into a rectangular shape approximately 2 x 3 mm to fit into the periodontal defect. In order to increase their hydrophilicity, the melt electrospun scaffolds were etched with 2M NaOH for o 30 minutes at 37 C followed by 5 rinses in ultrapure water. The scaffolds were sterilized in ethanol for 30 minutes followed by another 30 minutes of UV irradiation. These discs were utilized as a support membrane in order to harvest the cell sheet and facilitate their handling.

Covering barrier membrane

A solution electrospun PCL sheet was used as a barrier to cover the prepared periodontal defect following the insertion of the construct. This membrane was used to act as a barrier membrane to exclude the overlying connective tissues from the defect site during the healing phase. Membrane fabrication has been previously described (Vaquette and Cooper- White, 2011).

Briefly, the polymer was dissolved in a solution of chloroform and dimethylformamide (9:1, v/v) at a concentration of 15% w/v, then the solution was loaded into a 10 mL plastic syringe and electrospun at a feed rate of 2 mL/h, at 10 kV for 30 min with the flat collector surface placed at 20 cm from the syringe tip. Fiber diameter was 3 mm, 5 – 10 µm pore size and thickness of approximately 300-400 µm so this membrane behaved as a partially occlusive barrier membrane.

The physical and biomechanical integrity of the membrane was not compromised by being wetted and it retained its superior handling properties with excellent attachment and stability once positioned on top of the defect. The membrane was cut into smaller portions of 8mm x 5 mm dimensions.

Primary human periodontal ligament cells (h-pdl cells) isolation and proliferation

Primary h-pdl cells stored frozen in liquid nitrogen between second and third passages as described previously (Farag et al., 2014) were used in the study. Ethical approval for the collection of these cells was attained through the Griffith University Human Research Ethics Committee (DOH/17/7/HREC). The cells were revived and allowed to reach confluence for 2 weeks and passaged using 0.05% Trypsin and expanded into 175 cm2 flasks and expanded further for another week.

Cell sheet harvesting for in vitro experiment

Cells between 4th and 5th passages were seeded in 24 well plates with a seeding density of 5x104 cells/well in DMEM with 10% FBS. For the first 48 hr, ascorbic acid at a concentration of 1000μg/mL was added to enhance extracellular matrix formation (Beacham

et al., 2007). The cells were then grown for 19 days in media supplemented with one tenth the initial ascorbic acid concentration (100μg/mL); the culture media was changed every 48 hrs.

At the end of the 21 days of culture the cells had deposited sufficient ECM in order to enable the handling of the cell sheet. To collect the cell sheet, culture media was removed from each well and a PCL scaffold was carefully positioned in the center. The edges of the cell sheet were then gently detached from the plate and folded over the scaffold using sterile tweezers. The harvested cell sheets were placed in expansion media for 24 hours facing upward to allow adhesion onto the PCL scaffolds.

Perfusion Decellularization

The perfusion system used to decellularize the cell sheet constructs has been described previously (figure 3.1). An optimized technique of decellularization has been used in this study to decellularize the cell sheets (Farag et al., 2014). Briefly, the scaffolds were rinsed once with warm phosphate buffered saline (PBS) at 37oC and then placed in the decellularization chambers. Scaffolds were perfused in 30 mL of

20 mM NH4OH solution with 0.5% v/v Triton X-100. Bi-directional perfusion through the scaffolds was performed for 60 min at a rate of 1,000 mL/hr, with a flow inversion every 50 seconds. A DNase step involving the perfusion of a DNase I solution (100 U/mL, Invitrogen, o Cat. No. 18047-019) in CaCl2 (0.9 mM) and MgCl2 (0.5 mM) in sterile PBS at 37 C for 60 min and finally perfused with sterile water at 37oC for another 60 min then kept in PBS overnight at 4oC prior to subsequent utilization.

Recellularization of decellularized sheets for gene expression and STRO-1 immunostaining

Decellularized cell sheets were divided into two groups seeded with two different cell types, one of which was allogenic primary human periodontal ligament cells collected and cultured in a similar pattern to the PDL cells utilized for cell sheet fabrication. The other cell type was primary human placental mesenchymal stem cells (hp-MSC). The mechanism by which the MSC cells were obtained is described in a recent optimized technique performed by one of the research groups who provided these cells for the current study (Patel et al., 2014).

Cells were carefully seeded on top of the decellularized sheets with seeding density of 50 x 104 cells per sheet, sheets were wetted 3 times for 10 minutes with 20µl of DMEM to ensure cell adhesion to the sheets. After 30 minutes post seeding, constructs were completely covered in nutrition medium that was changed every 48 hrs. Samples were collected after 3, 7 and 14 days, rinsed with PBS and transferred to new wells in another plate. Some samples were kept at -80ºC for later RNA extraction and the others were fixed in 4% paraformaldehyde and kept at 4ºC for subsequent confocal imaging.

RNA extraction and cDNA synthesis

For gene expression analysis, six samples were pooled together for RNA extraction in 1 ml of Trizol (Life technologies). RNA was extracted according to the manufacturer’s instructions and quantified by spectroscopy using a NanoDrop (Thermo Scientific). RNA was converted to cDNA using M-MuLV reverse transcriptase (New England Biolabs) following the manufacturers instruction.

Briefly 1 µg of RNA was reversed transcribed in a two-step reaction which consisted of an initial denaturation and priming reaction (70ºC and 4ºC respectively) containing 1 µL oligo dT primer (40 µM), 1 µL random hexamer (40 µM), 1 µL dNTP (10 µM), 1 µg RNA, and nuclease free water to 16 µL. Subsequently, 1 µL of M-MuLV reverse transcriptase (200 units / µL) and 2 µL of 10x RT buffer were added to the reaction. cDNA synthesis was carried out at 40ºC for 1 hour followed by enzyme deactivation at 90ºC for 10 min. The cDNA was diluted 3 fold and stored at -20ºC until further use.

Gene expression analysis

PCR primers of the genes analysed in this study are listed in table 4.1. The 10 µL qPCR reaction consisted of 5 µL KAPA SYBR FAST Roche LightCycler 480 mix, 0.2 µL of each primer (10 µM), 2 µL of cDNA, 2.6 µL of nuclease free water. The reaction was run on a Roche LightCycler 480 according to the conditions specific to the SYBR mix and CT values were obtained. Analysis of the gene expression data was carried out according to the 2- ∆∆CT method. Cells seeded on PCL scaffolds alone were used as controls.

Human stromal precursor antigen-1 (STRO-1) immunostaining and confocal imaging

Primary human placental mesenchymal stem cells (hp-MSC) were seeded on melt e- spun PCL scaffolds and decellularized human periodontal ligament cell sheets with seeding density of 5 x 104 cells per sheet/scaffold. Samples were collected at 3, 7 and 14 days respectively for immunostaining. Briefly, samples were fixed with a 4% paraformaldehyde solution at pH 7.4 (Sigma- Aldrich, Australia) in phosphate buffered saline (PBS) for 20 minutes and thereafter rinsed with PBS. The samples were then incubated for 30 minutes in bovine serum albumin (BSA 1%, Sigma-Aldrich, Australia) in PBS to block non-specific binding of the primary antibody. Aliquots of 100 µL of mouse anti-human STRO-1 monoclonal antibody were used at a dilution of 1:100 in 1% BSA in PBS for 1 hour at room temperature. Samples were rinsed with PBS then incubated in a solution of BSA (1%) in PBS containing the goat anti-mouse IgG1-fluorescein isothiocyanate (FITC - sc-2078, Santa Cruz Biotechnology) secondary antibody diluted 1:200 for another 1 hour in the dark. Confocal imaging was done with a Nikon A1R+ microscope (Griffith Health Centre, Griffith University) with excitation wavelength of 490 nm and emission was collected at 520 nm.

In vivo study

Cell sheet preparation and decellularization

Human periodontal ligament cells between 3rd and the 5th passages were seeded in 6 well plates at a seeding density of 5 x 104 cells/well and cultured in DMEM nutrition media supplemented with 1000 µg/ml ascorbate-2-phosphate (Sigma-Aldrich) at 37ºC for 48 hrs.

Table 4.1: List of primers for gene expression study.

Gene Abbreviation Primer Name Sequence Tenascin-C TNC TNCHF ATGTCCTCCTGACAGCCGAGAA

TNCHR AGTCACGGTGAGGTTTTCCAGC

Osteopontin OPN OPNHF CGAGGTGATAGTGTGGTTTATGG

OPNHR GCACCATTCAACTCCTCGCTTTC

Osteocalcin OCN OCNHF CGCTACCTGTATCAATGGCTGG

ONCHR CTCCTGAAAGCCGATGTGGTCA

Collagen Type1 – A1 COL1A1 COL1A1HF GATTCCCTGGACCTAAAGGTGC

COL1A1HR AGCCTCTCCATCTTTGCCAGCA

Collagen Type1 – A2 COL1A2 COL1A2HF CCTGGTGCTAAAGGAGAAAGAGG

COL1A2HR ATCACCACGACTTCCAGCAGGA

Collagen Type3 – A1 COL3A1 COL3A1HF TGGTCTGCAAGGAATGCCTGGA

COL3A1HR TCTTTCCCTGGGACACCATCAG

Alkaline phosphatase ALPL ALPLHF GCTGTAAGGACATCGCCTACCA

ALPLHR CCTGGCTTTCTCGTCACTCTCA

Vascular endothelial growth factor VEGF VEGFHF TTGCCTTGCTGCTCTACCTCCA

VEGFHR GATGGCAGTAGCTGCGCTGATA

Glyceraldehyde 3-phosphate GAPDH GAPDHF GTCTCCTCTGACTTCAACAGCG dehydrogenase GAPDHR ACCACCCTGTTGCTGTAGCCAA

Beta-actin ACTB ACTBHF CACCATTGGCAATGAGCGGTTC

ACTBHR AGGTCTTTGCGGATGTCCACGT

Bone morphogenetic protein BMP2 BMP2HF TGTATCGCAGGCACTCAGGTCA

BMP2HR CCACTCGTTTCTGGTAGTTCTTC

Culture media was supplemented with 100 µg/ml ascorbate-2-phosphate for another 18 days with changing of media every 48 hrs. After the 21st day of culture, cell sheets were gently detached from the borders of each well as described before and folded four times to conform to the size of the periodontal defect in the rat mandible. Medical grade PCL scaffolds were cut into 3 mm x 1.5 mm rectangles and utilized to harvest the folded cell sheets. All samples were left overnight in DMEM at 37 ºC to insure strong physical attachment between cell sheets and the carrier PCL scaffolds. Some samples were decellularized using the perfusion technique previously mentioned in this study where five constructs were decellularized at a time and then transferred into sterile PBS solution and were incubated at 37 ºC to be placed inside the prepared periodontal defect afterwards (figure 4.1B).

Rat periodontal defect

The periodontal defect model was adopted from a previous study (Dan et al., 2014). The study protocol was approved by the Animal Ethics Committee of Griffith University (DOH/02/11/AEC). Twelve week old athymic rats (Rattus norvegicus, Strain-CBH-rnu/Arc, Animal Resources Centre, Murdoch, Western Australia) were carefully placed in a gas chamber and anaesthetized by inhalation, using Isoflurane (AttaneTM, Bomac Animal Health Pty Ltd, Australia).

The surgical site areas were shaved and disinfected using gauze swabs wetted with Povidone-Iodine, then a full thickness incision was made along the skin of the inferior border of the mandible. The masseter muscle and periosteum were reflected as to expose the buccal cortical bone of the mandible. The covering alveolar bone and part of root cementum of the mandibular first molar were removed using round burs with copious saline irrigation. The surgical defect was made approximately 1.5 mm height x 3 mm width and 1 mm in depth (figure 4.2A).

Mandibular defects that received constructs with decellularized sheets were assigned as the test group; defects receiving constructs with fresh periodontal ligament cell sheets and plain PCL scaffolds were used as controls. Constructs of both decellularized and fresh cell sheets where placed so the sheet was facing the exposed roots and made to fit precisely into defect site (figure 4.2B). Solution electrospun medical grade PCL membrane was used to

cover the area of defect to exclude connective tissue from the defect site (figure 4.2C). The masseter muscle was repositioned and approximated with 5-0 bio-absorbable sutures (VICRYL RAPIDE™ (polyglactin 910), Ethicon), and covering skin was approximated and sutured with 5-0 Polypropylene Suture (PROLENE®, Ethicon) to ensure healing with primary intention.

To control post-surgical pain, Temgesic® (0.02 mg/kg) was used as analgesic. Prophylactic antibiotics were administrated as an antibacterial oral solution (200 ml/day of Enrofloxacin (Baytril®25), active ingredient 25 mg/ml Enrofloxacin, Bayer Australia Ltd) each day postoperatively. Rats were sacrificed at 2 and 4 weeks post-surgery and the mandible samples (n = 5 for each group at each time point) were collected (figure 4.2D&E) and fixed in 4% paraformaldehyde (PFA) solution overnight at room temperature and then rinsed with PBS.

Figure 4.1: Decellularized construct fabrication. (A) PCL scaffold. (B) Decellularized constructs. (C&D) Perfusion system and decellularization chambers.

Figure 4.2: Surgical defect preparation with decellularized construct utilization (A) Rat mandibular surgical defect prepared (arrow pointing at defect). (B) Arrow pointing at decellularized construct properly positioned inside the bony defect. (C) Covering PCL membrane placed on top of the construct to secure its position. (G&E) Rat mandible post- sacrifice with integrated construct.

Microcomputed tomography (µCT)

A microcomputed tomography (micro-CT) scanner (mCT40, SCANCO Medical AG, Brüttisellen, Switzerland) was used to scan rat mandibles at a resolution of 12 µm, a voltage of 45 kVp and a current of 177 mA. Three-dimensional (3D) reconstructed images were prepared using the micro-CT scanner software.

Hematoxylin and eosin (H&E) and Mason Trichome histology (MT) staining

Samples were decalcified in 15% Ethylenediaminetetraacetic acid (EDTA) solution with regular change of solution every 72 hours for the first 2 weeks, then once a week for another 6 weeks. Samples were paraffin embedded and sectioned horizontally every 4-5 µm with a slide stained every 10 sections to locate the defect. Samples were stained with both H&E solutions and Mason Trichome staining kit according to manufacturer instructions.

For Mason Trichome staining; briefly, samples were deparaffinized and rehydrated starting with 100% absolute alcohol then through a series of alcohol dilutions ending with 50% alcohol. Samples were washed three times in distilled water. Samples were re-fixed in Bouin’s solution for 60 minutes at 56ºC to improve staining quality. Samples were rinsed under running water for 10 minutes then stained in Wiegert’s iron hematoxylin working solution for total of 10 minutes then subsequently washed with warm tap water followed by distilled water at room temperature. Samples were stained in Biebrich scarlet-acid fuchsin solution for 15 minutes then washed for 5 minutes. Tissues were differentiated in phosphomolybdic/phosphotungstic acid solution for 10-15 minutes then transferred into aniline blue solution for another 10 minutes, rinsed in distilled water and differentiated in 1% acetic acid solution for 5 minutes. Samples were washed in distilled water then dehydrated through a serious of alcohol dilutions ending with absolute alcohol then placed in xylene solution before mounting.

Immunohistochemistry

Sections were affixed to Menzel Superfrost Plus adhesive slides and air-dried overnight at 37ºC then dewaxed and rehydrated through descending graded alcohol concentrations to PBS pH 7.4, using standard protocols. Antigen retrieval was done using trypsin enzyme 0.05% pH 7.8 for 10 minutes. Sections were washed in three changes of PBS for 5 minutes each. Endogenous peroxide was removed by incubating in 3% hydrogen peroxide solution for 15 minutes. Non-specific antibody binding was inhibited by incubating the sections in Biocare Medical Background Sniper (BS966 – Biocare Medical) for 30 min. Excess normal serum was decanted from the sections and the primary antibodies for osteopontin and osteocalcin (OPN AKm2A1 SC 21742, OSTCAL FL-05 SC30045 - Santa Cruz) were diluted at 1/25 and

1/50 respectively in PBS, applied on tissue sections then incubated overnight at room temperature. Sections were washed in three changes of PBS for 5 minutes then incubated with Biocare Medical Mouse probe for 15 minutes at room temperature. Samples were incubated with Biocare Medical Universal HRP polymer (M4U534– Biocare Medical) for 15 minutes at room temperature. Sections were washed in three changes of PBS for 5 minutes each. Color was developed in Nova Red for 5 minutes and was checked microscopically. Finally sections were washed in water then counterstained in hematoxylin.

Statistical analysis

Results were expressed as mean± standard error of mean from at least five biological replicates and a one-way ANOVA test was used to analyze the data. The significance level for the statistical analysis was set at p < 0.05.

Results

Decellularised cell sheets induce bone markers gene expression in human placental mesenchymal stem cells and periodontal ligament cells

A summary of the gene expression changes in human placental mesenchymal stem cells and periodontal ligament cells (PDL) following seeding on decellularized sheets compared to the PCL scaffold control at different time points is summarized in Table 4.2.

Gene expressions of human periodontal ligament cells

Alkaline Phosphatase enzyme (ALPL)

PDL showed expression of ALPL only on the 14th day after seeding. Expression of ALPL on decellularized sheets was equivalent to that of the control PCL samples (figure 4.2A).

Bone Morphogenetic Protein type 2 (BMP-2)

BMP2 gene was expressed by PDL cells only on the 14th day where it was upregulated three folds on decellularized PDL sheets (figure 4.3B).

Osteocalcin (OCN)

OCN gene was expressed on the 14th day by PDL cells showing two-fold upregulation on decellularized sheets (figure 4.3C).

Osteopontin (OPN)

In response to the decellularized cell sheets, OPN gene was significantly upregulated at day 14 compared to days 3 and 7. Furthermore, on day 14, OPN was upregulated by 11 fold on the decellularized sheets compared to the control PCL scaffold (4.3E).

Vascular Endothelial Growth Factor (VEGF)

VEGF was only detected on the 14th day where the gene expression was upregulated almost three fold in PDL cells on the decellularized sheets (figure 4.3D).

Tenascin C (TNC)

TNC was upregulated in PDL cells on the 3rd and 7th day by one and a half fold, then downregulated on the 14th day (figure 4.3F).

Human collagen type 1A1 (COL1A1)

COL1A1 was upregulated in PDL cells on the third day after seeding by almost two fold, then becoming equivalent to that expressed on the control PCL scaffold on the 7th day, and was eventually downregulated on the 14th day (figure 4.3G).

Human collagen type 1A2 (COL1A2)

COL1A2 was upregulated by two fold on the third day and increased to four fold on the 7th day. However on the 14th day it became equivalent to that expressed by PDL cells seeded on the control PCL scaffold (figure 4.3H).

Human collagen type 3A1 (COL3A1)

COL3A1 was downregulated on the 14th day after seeding compared to the control PCL scaffold (figure 4.3I).

Gene expression of human placental mesenchymal stem cells

Only genes related to collagen markers were expressed by the placental mesenchymal stem cells.

Human collagen type 1A1 (COL1A1)

COL1A1 was significantly downregulated at the first two time points, then by day 14, the cells seeded on the decellularized sheets induced statistically significant gene upregulation of approximately one and a half fold compared to that expressed by cells seeded on PCL scaffolds (figure 4.3J).

Human collagen type 1A2 (COL1A2)

COL1A2 expression was unchanged compared to the control at all time points. (figure 4.3K).

Human collagen type 3A1 (COL3A1)

COL3A1 expression was expressed at all three time points but had no significant change in expression compared to the control (figure 4.3L).

Figure 4.3: Graph showing different gene expressions in HPDL and human placental mesenchymal cells after seeding on decellularized matrix compared to PCL scaffold as a control at three time points. 85

Different gene expressions of markers relevant to bone and periodontal ligament by HPDL cells (A-I). Gene expressions of collagen markers by human mesenchymal cells (J-L).

Table 4.2: Summary of gene expression results.

3 DAYS 7 DAYS 14 DAYS

GENES HPDL HPC HPDL HPC HPDL HPC

ALPL ND ND ND ND N ND

BMP2 ND ND ND ND ++ ND

OCN ND ND ND ND + ND

OPN - ND - ND ++ ND

VEGF ND ND ND ND ++ ND

TNC + ND + ND + ND

COL1A1 + - N - - +

COL1A2 + N ++ + N N

COL3A1 ND N ND N - N HPDL = HUMAN PERIODONTAL LIGAMENT CELLS, HPC = HUMAN PLACENTAL CELLS, ND = NON DETECTED, N = NEUTRAL, (+) = UPREGULATION, (++) = upregulation ≥ 2 folds, (-) = DOWNREGULATION

Differentiation of placental mesenchymal stem cells initiated by the decellularized cell sheets shown by loss of STRO-1 expression

Immunostaining against STRO-1, which is a marker for undifferentiated mesenchymal stem cells (Lin et al., 2011; Ning et al., 2011), was detected by confocal imaging in placental mesenchymal stem cells seeded on top of control PCL scaffolds and decellularized periodontal ligament sheet constructs after 3 days of seeding.

The strong STRO-1 expression remained on the 7th and 14th days post seeding on the control PCL scaffolds indicating the maintenance of an undifferentiated phenotype (figure 4.4A, 4.4B &4.4C). However, STRO-1 expression on the decellularized sheet constructs diminished by the 7th day (figure 4.3E) reaching its lowest expression by day 14 post seeding (figure 4.4F).

Figure 4.4: Stro-1 expression in human placental stem cells on PCL scaffold and decellularized periodontal ligament cell sheet 3, 7 and 14 days’ post seeding. Strong

STRO-1 expression remained over 14 days post seeding on PCL scaffolds (A-C). STRO-1 expression on decellularized sheets diminished by the 7th day (E) day 14 post seeding (F).

In vivo evaluation of the decellularized cell sheets in a rat mandibular defect

Micro CT

Three dimensional reconstructed images of rat mandibular defects that received PCL scaffolds, decellularized constructs and fresh periodontal ligament cell sheets respectively, did not show any significant bone fill in any of the aforementioned groups after two weeks post-surgery. Minimal bone fill is seen only at the boundaries of the defects (figure 4.5 A, B & C). After 4 weeks post-surgery, more bone fill was detected at the defects’ boundaries (figure 4.5 D, E & F).

Histology and immunohistochemistry

Histology sections of periodontal defects that received the decellularized constructs showed evidence of biological compatibility and integration between the decellularized sheets and the surrounding periodontal tissues at both time points. The decellularized constructs were completely infiltrated by host cells, demonstrating a high potential for in vivo recellularization (figure 4.6 e-h and figure 4.7 e-h). These histological outcomes were comparable to those observed in the defects that received fresh periodontal ligament cell sheets, where both were well integrated into the surrounding tissues with no signs of adverse tissue reaction (figure 4.6 i, j, k & l). Noticeably, the defects that received the decellularized constructs showed dense fibrous tissues facing the surface of roots (figure 4.6h).

Samples that received empty PCL scaffolds showed cell infiltration into the scaffolds where cells were detected lining the empty spaces previously occupied by the PCL fibers (Figure 4.6c & 4.6d). Empty spaces representing PCL fibres were found to be much closer to root surfaces compared to the other groups, due to the absence of cell sheets which also have a space maintain role (Figure 4.6d). None of groups showed significant new bone formation except for limited amounts at the defect boundaries as seen in both figure 4.6 and 4.7.

At the 2 week’ time point, the orientation of collagen fibres adjacent to the exposed tooth roots within the defects in all groups was found to be unorganized with no clear direction and hence these collagen fibres could not be classified as newly formed periodontal ligament (figure 4.7 b, f & j).

Collagen fibres adjacent to root surfaces became more organized by the four week time point. In the defects that received PCL scaffold only, fibre orientation was almost parallel to the root surface with no obvious fibres inserted into the cementum surface (figure 4.7d). However, in the defects that received decellularized constructs, Collagen fibres between the PCL fibre spaces (that appeared as aggregation of white circles as seen in histological sections) and root surfaces in the defect area were more organized and some fibres were inserted perpendicularly into newly formed cementum (figure 4.7h). Samples that received the fresh cell sheet showed the most organized perpendicular fibre orientation into root surface cementum and higher cellularity when compared to the other groups (figure 4.7 l).

Immunostaining of histological sections for osteocalcin (OCN) and osteopontin (OPN) showed positive staining of the original native bone and teeth hard tissues that were not exposed to the defect areas, Also OCN and OPN were expressed in the periodontal ligament cells distant from the defect area (figure 4.8 and 4.9).

Defects that received plain PCL constructs did not show positive OCN staining on the root surfaces or within the scaffold at both the 2 and 4 week time points (figure 4.8b & 4.8d). Defects that received both decellularized and fresh periodontal ligament cell sheet constructs showed positive OCN staining at both time points between the constructs and the root surfaces; whereby the defects that received fresh cell sheets showed stronger staining at both time points when compared to the decellularized constructs (figure 4.8f, h, j &l) indicating higher mineralization activity.

OPN staining was positive in all groups; however, OPN expression was stronger in defects at the second time point (four weeks post-surgery). The OPN expression was found to localize mainly around the PCL fibre spaces and at the margins of the bony defect, but it was not significantly expressed on the root surfaces exposed to the defect areas (figure 4.9).

Figure 4.5: micro CT 3D reconstructed image data showing created mandibular defects in rat mandibles treated with PCL scaffold, decellularized construct and fresh cell sheet with different degrees of bone fill at margins of the defects. PCL scaffold as control showed almost no bone formation within the defects (A&D), decellularized constructs and fresh cell sheet constructs showed some bone formation detected mainly at the defect margins (B&E and C&F respectively).

Figure 4.6: Histological H&E stained sections of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery. Images to the right of boxes are magnified versions. No signs of new bone formation in defects treated with PCL (control) with haphazardly distributed periodontal ligament fibers (a-d). More organized periodontal ligament fibers with signs of new bone formation were seen in defects treated with decellularized constructs (e-h) and fresh cell sheets constructs (i-l).

Figure 4.7: Histological Mason Trichome stained sections of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery. Images to the right of boxes are magnified versions. No significant collagen fibers insertion into root surfaces as seen in control group over 4 weeks (a-d). Well organized periodontal ligament fibers inserted into cementum were seen between scaffold fibers and root surfaces in defects treated with both decellularized constructs

(e-h) and fresh cell sheets constructs (i-l) at week 4 post-surgery. ( ) Symbol refers to the acute angle formed by the insertion of newly formed periodontal ligament into newly formed cementum. ∢

Figure 4.8: Osteocalcin (OCN) immunohistochemistry staining of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery. OCN immunostaining was not detected in control group (a-d), but evident in defects treated with decellularized constructs and fresh cell sheet constructs on both 2 and 4 weeks post- surgery (e-h) & (i-l) being more significant in the later. Arrows point at areas considered positively stained (brown colour).

Figure 4.9: Osteopontin (OPN) immunohistochemistry staining of rat mandibular defects that received PCL scaffold, fresh cell sheet as controls and decellularized construct as the test group at two and four weeks post-surgery. OPN immunostaining was detected in control group (a-d), but more evident in defects treated with decellularized constructs and fresh cell sheet constructs on both 2 and 4 weeks post-surgery (e-h) & (i-l) being more significant in the later. Arrows point at areas considered positively stained (brown colour).

Discussion

Decellularized extracellular matrices have been gaining more attention as an approach to fabricate biological scaffolds (Bloch et al., 2011; Cho et al., 2005; Farag et al., 2014). In this study, novel tissue engineered decellularized constructs were assessed for their regenerative potential in periodontal defects. The decellularization method has been previously optimized in an earlier study (Farag et al., 2014) where more than 97% of the cellular contents of the cell sheets were removed, whilst maintaining the native structural and functional properties.

Real time PCR was performed in this study to assess the gene expression of bone and periodontal tissue markers by cells seeded on decellularized periodontal ligament sheets compared to control empty PCL scaffolds. Allogenic human periodontal ligament cells and human placental mesenchymal stem cells were used to assess the influence of the cell sheets on the differentiation of these multipotent progenitor cells (Castrechini et al., 2010; Wulf et al., 2004).

The gene expression markers were selected for their relevance to periodontal wound healing and regeneration, and encoded for proteins localized in bone, cementum and periodontal ligament tissues (table 1). Human periodontal ligament cells from an allogenic source generally expressed higher levels of bone markers when seeded on decellularized periodontal ligament sheets (figure 4.3 B, C & E) as demonstrated by increased BMP2, Osteocalcin and Osteopontin expression compared. Given that these are expressed in native human periodontal tissues (Ivanovski et al., 2001; Zhu and Liang, 2015), upregulation of these marker genes is supportive of the notion that the decellularized sheets can promote the differentiation of repopulating progenitor cells.

It is noteworthy that ALPL, BMP2, OCN, OPN and VEGF could not be detected in the first two time points at 3 and 7 days respectively. This could be explained by an initial proliferative response by the of the repopulating PDL cells before differentiation is induced. This is consistent with the rapid proliferation of allogenic PDL cells on the decellularized matrices reported in Chapter 2 (Farag et al., 2014). In contrast, the collagen markers COL1A1

and COL1A2 were upregulated at the earlier two time points of 3 and 7 days respectively then downregulated by the 14th day (figure 4.3G & 4.2H), which is likely related to enhanced extracellular matrix production.

The upregulation of VEGF expression on the 14th day by PDL cells seeded on decellularized PDL sheets suggests a pro-osteogenic effect, which might subsequently improve the healing and regenerative capacity of these decellularized constructs when used in vivo (figure 4.6h).

Only COL1A1 & A2 and COL3A1 genes were expressed by the human placental mesenchymal stem cells at all time points. When compared to the control PCL scaffold, The placental cells seeded on decellularized matrices showed downregulation of the COL1A1 gene in the first two time points followed by slight upregulation on the 14th day (figure 4.3J). A possible explanation is that these immature cells are initially stimulated to proliferate by the retained growth factors, such as FGF, HGF and VEGF, which were shown to be retained in the decellularized sheets in chapters 2 and 3 (Farag et al., 2014). By the 14th day, as suggested by the STRO-1 expression results, these mesenchymal stem undergo differentiation to a more anabolic cell line that is likely to produce increased levels of extracellular matrix, as reflected by upregulated COL1A1 expression.

The suggestion of accelerated mesenchymal stem cell differentiation on the decellularized constructs was further supported by the temporal decrease in immunostaining against the ‘stemness’ marker STRO-1 (Castrechini et al., 2010). The expression of this marker remained relatively strong over the three time points when seeded on the control PCL scaffolds (figure 4.4 A, B & C) but the staining was reduced at each time point after seeding on decellularized sheets (figure 4.4 D, E & F).

The gene expression changes in undifferentiated stem cells following recellularization of a variety of decellularized human/animal tissues or organs, such as tendon, bone and dermis (Yin et al., 2013), kidney (Nakayama et al., 2013) and cartilage (Sutherland et al., 2015). However, few studies have assessed gene expression and cell differentiation on tissue engineered cell derived matrices. One study evaluated the ability of adult human mesenchymal stem cell (hMSC)-derived ECM to rescue the phenotype of osteoarthritic chondrocytes and to further stimulate the differentiation of healthy chondrocytes (Thakkar et al., 2013). A second study used decellularized cell derived matrix laid down by both MC3T3- E1 mouse preosteoblasts and NIH3T3 mouse fibroblasts to evaluate human mesenchymal

100

stem cell differentiation and gene expression of markers such as Osteopontin, Collagen type I&II and Aggrecan (Choi et al., 2014). The difference between the aforementioned two studies and the present study is that instead of using osteogenic and chondrogenic media to induce cell differentiation, this study relied solely on the native structural and biological characteristics of the decellularized sheets to induce cell differentiation and bone markers.

In the in vivo part of this study, the main aim was to evaluate the biocompatibility and integration of decellularized periodontal ligament cell sheet constructs and compare it to plain PCL scaffold and fresh periodontal ligament cell sheet when introduced into surgically created periodontal defects in nude rats. In a previous study, fresh periodontal ligament cell sheets were shown to enhance periodontal attachment formation and regeneration these type of defects (Dan et al., 2014). The decellularized constructs were compared to both a positive control of fresh PDL sheet constructs and a negative control of PCL scaffold alone.

It was found that the constructs in all three groups were integrated into the surrounding tissues, with host cell infiltration into the constructs and no signs of tissue rejection (figure 4.6). The decellularized constructs did not appear to negatively influence periodontal attachment formation compared with the positive control of fresh cell sheet constructs, while both decellularized and fresh cell sheet constructs performed better than the empty scaffold. These findings support the clinical potential of the decellularized constructs, whereby the bioactivity of the retained extracellular matrix can be harnessed without the potentially negative immunogenic effects of the cellular material.

In this study we could not detect significant amounts of new bone formation in any of the groups either by micro CT (figure 4) or histologically (figure 4.6 & 4.7), which is in contrast to a previous study that compared fresh cell sheets from different intraoral sources (Dan et al., 2014). This could be attributed to the use of osteogenic calcium phosphate coating in the previous study, which was not utilized in this study in order to minimize the variables that could influence wound healing.

Although obvious bone formation was not induced in the periodontal defects by any of the groups at the two time points, positive OCN and OPN staining was expressed within the PCL scaffolds and between the scaffold and root surfaces within the defects (figure 4.8 & 4.9). This positive expression of these two bone markers were more prominent in the defects that received the decellularized and fresh periodontal ligament cell sheet constructs, mainly at the 4 week time point. The positive OPN and OCN expression in the defects receiving

101

decellularized periodontal ligament cell sheets is an indication that these matrices have the potential to initiate remineralisation inside the periodontal defects, and supports the findings of upregulated OPN and OCN gene expression in the in vitro study. This result is also consistent with the immunohistochemical findings in the previous study (Dan et al., 2014) although the expression was less intense, perhaps due to the use of plain rather than calcium phosphate coated scaffold in the current study.

In conclusion, decellularized periodontal ligament cell sheet constructs promoted the differentiation of PDL and mesenchymal stem cells. In particular, PDL cells seeded on top of decellularized matrices showed upregulation in the expression of mineralized tissue markers, especially after two weeks post-seeding, when compared to PCL scaffolds alone. The decellularized cell sheet constructs were shown to support periodontal attachment in a rat periodontal defect model, and hence further investigation is warranted to explore the possible clinical utilization of this approach.

102

Chapter 5

Discussion and future directions

103

The periodontal defect resulting from periodontitis is generally very difficult to regenerate due to the complex structure of the periodontium (Armitage et al., 2003; Cho and Garant, 2000; Nanci and Bosshardt, 2006a).

Of the current possible treatment options, none of them are capable of fully restoring the lost periodontal tissues; instead very limited areas of true regeneration confined usually to the most apical part of the defect is achieved, with the rest of the defect area healing via reparative (scarring) tissue (Giannobile and Somerman, 2003; Polimeni et al., 2006; Sculean et al., 2003; Sculean et al., 2004a; Yukna and Mellonig, 2000).

Tissue engineering is considered to be a promising approach for periodontal regeneration as it can allow greater control of the periodontal wound healing process by orchestrating the interaction between the various components of the periodontium. Many research groups have studied different tissue engineering approaches using combinations of bioactive molecules, such as bone morphogenetic proteins and enamel matrix derivatives, combined with host cells grown in vitro to be re-implanted again at the defect site with variable degrees of success (Bartold et al., 2000; Danesh-Meyer, 2000; Izumi et al., 2011; Jin et al., 2003; Kaigler et al., 2006; Kao et al., 2009; Mao et al., 2006).

A Japanese (Okano’s) research group were the innovators of a new technology using temperature responsive culture plates to harvest intact mature cells sheets. This method has been used to obtain periodontal cell sheets that have been implanted into periodontal defects to promote regeneration (Akizuki et al., 2005; Flores et al., 2008a; Flores et al., 2008b; Ishikawa et al., 2009). Our laboratory has utilised this new cell sheet technology to investigate cell lines from various intraoral sources, namely alveolar bone, gingiva and periodontal ligament (Dan et al., 2014) and found that periodontal ligament cell sheets performed the best in terms of simultaneous regeneration of root cementum, alveolar bone and functional periodontal ligament fibers.

The use of cell sheets in regenerative medicine require certain essential conditions and facilities. Since the cells need to be harvested from the host for in vitro culture before reimplantation, both clinical and cell culture facilities are required and should ideally be co- located. Therefore, aside from clinical expertise, there is also a requirement for cell biology capability for the duration of the cell sheet culture period, with obvious cost implications. With the approach utilized in this study, our goal was to develop a cost-effective cell sheet strategy that could ultimately be utilized clinically as an “off-the-shelf” product.

104

Similar to cell sheet research, the field of decellularized matrices has also been the subject of significant recent attention. This is due to the fact that decellularized matrices obtained from tissues, organs and cell derived extracellular matrices (ECM) have been shown to biologically and functionally mimic the native host tissues and have subsequently been used as biomaterials for regenerative purposes (Badylak, 2007; Badylak et al., 2009; Gilbert et al., 2006). In order to optimize the possibility of a successful outcome by utilizing multi- disciplinary expertise, this study combined the two technologies of cell sheet fabrication and decellularization to develop a cell derived matrix that has periodontal regenerative potential.

Because obtaining a decellularized cell sheet was a very novel approach that we aimed to optimize, it was essential to standardize several parameters. Firstly, human periodontal ligament cells (HPDLC) were chosen as a primary cell source, as it was easy to obtain and had been shown to promote periodontal regeneration in previous experiments (Vaquette et al., 2012; Vaquette et al., 2013). Secondly, it was decided to harvest the cell sheets manually using fine sterile tweezers instead of the thermal responsive plates utilized in earlier studies (Dan et al., 2014), as this method was found to be easier and quicker, as well as being more cost effective (see chapter 2).

Initially, a decellularized extracellular matrix formed by HPDLC in vitro was obtained using a simple technique to remove the cells from a monolayer (Beacham et al.,

2007). This technique relied on chemical decellularization with NH4OH and Triton X-100. After successfully identifying collagen type I and fibronectin as the main components of the ECM formed by HPDLCs, the method of decellularization was modified for use in mature cell sheets in terms of the concentrations of chemicals and the DNase used. Also we designed a simple perfusion devise to assist with the decellularization procedure (see chapter 2).

In the first study (chapter 2) decellularized sheets were shown to retain most of their structural architecture and preserved their ECM collagen contents as well. Growth factor (basic-FGF, HGF and VEGF) retention was also assessed, and it was found that they were reduced to 1/10th of their original concentrations but were nevertheless much higher than levels seen in a similar study (Reichert et al., 2010). Cell proliferation on top of the decellularized sheets was tested using allogenic HPDLCs and were shown to have a faster rate of growth when compared to cell growth on commercially available collagen (BioGide®, Geistlich, Switzerland) membranes.

105

Melt electro-spun polycaprolactone (PCL) was chosen to fabricate scaffolds that were used to support the decellularized cell sheets, and were utilized for all experiments performed in these studies. The advantage of using PCL is the relatively cheap price, ready availability, and established regulatory approval for clinical use. This material has been shown to be compatible with cells when used both in vitro and in vivo experiments in numerous studies (Ahn et al., 2015; Dan et al., 2014; Thakkar et al., 2013). The PCL scaffolds are an essential component of the decellularized cell sheet construct as they act as a carrier for the delicate cell sheets, providing support and ease of handling (see chapter 2 & 3). Random melt electrospun was used in the initial study (see chapter 2), as this was an easy and fast fabrication method requiring minimal technical skills. However, it is noteworthy that individual fiber thickness within the scaffolds was variable and difficult to standardise. This did not create any issues for our early experiments as the scaffold was able to readily fulfil its primary role of providing support and facilitating handling of the cell sheets.

In the second study the optimised decellularization protocol described in Chapter 2 was assessed against other decellularization protocols described in the literature. Therefore, 10 different decellularization protocols were tested in both static and dynamic conditions using different chemical and thermal approaches. It was found that the decellularization method optimised in the first study outperformed the other protocols in terms of structural integrity, DNA elimination and growth factor retention.

Apart from optimization of the decellularization protocol, the perfusion device was upgraded to insure optimum decellularization while maintaining a completely closed and sterile environment for the constructs.

Furthermore, the structure and architecture of the PCL scaffolds used as cell sheet carriers was optimized using a direct writing melt electrospinning technology which resulted in more organized and standardized constructs (see chapter 3).

A gene expression analysis was performed to assess the influence of decellularized cell sheets on allogenic HPDLCs, where undifferentiated human placental mesenchymal stem cells were used for comparison. We successfully showed the influence of decellularized sheets on seeded HPDLCs using a broadly representative panel of bone and periodontal ligament markers, and concluded that important genes for hard tissue formation such as osteocalcin, ostepontin and ALPL were upregulated by the decellularized constructs compared to control empty PCL scaffolds. Furthermore, gene expression of vascular

106

endothelial growth factor (VEGF), which is essential for endothelial cell growth and proliferation, was also upregulated on the decellularized constructs (see chapter 4).

It was essential to evaluate the compatibility and bio-integration of the decellularized sheets when implanted in vivo into a surgically prepared periodontal defect. A small animal model (rat) was chosen as the surgical protocol was previously optimized (Dan et al., 2014). The main aim of this in vivo experiment was to evaluate how the surrounding tissues of the prepared defect would integrate with the decellularized sheet. Based on histological assessment, we found that the decellularized constructs could integrate in a similar pattern as a fresh cell sheet.

Furthermore, new collagen fiber attachment to the root surfaces was observed in the decellularized construct group to a similar degree as the positive control (fresh PDL cell sheets) and superior to the negative control (empty scaffold). Since fresh PDL cell sheet constructs have been shown to strongly support periodontal regeneration in several models (Vaquette et al, 2012; Dan et al, 2014), the demonstration of equivalent outcomes by the decellularized constructs is very encouraging in terms of future clinical applicability.

Immunohistochemical analysis revealed positive staining to OCN and OPN similar to that seen with the fresh cell sheet constructs. We could not detect significant new bone formation within the periodontal defects, unlike a previous study conducted by our group where significant bone formation was noted (Dan et al, 2014). However, the previous study used calcium phosphate (CaP) coated scaffolds with enhanced osteoconductivity, while in the current study non-coated PCL scaffolds were used in order to evaluate the effect of the decellularized extracellular matrix without other confounding/modifying factors. In this context, the immunohistochemical evidence of OCN and OPN protein expression, along with the gene expression studies showing upregulated bone marker expression suggest that the decellularized constructs have the potential to support mineralized tissue formation, which in turn is essential for the formation of key periodontal structures, namely bone and cementum.

107

Conclusion

This study aimed to develop a novel approach to fabricate cell derived extracellular matrix in the form of a tissue engineered decellularized periodontal ligament cell sheet construct. The decellularized construct was found to retain its biological and architectural integrity, and showed good biocompatibility and the ability to support periodontal attachment in vivo. Taken together, the research findings presented in this thesis suggest that decellularized cell sheet constructs represent a promising candidate for periodontal regeneration. Furthermore, the removal of the potentially immunogenic cellular component makes these constructs an attractive candidate for a tissue engineered ‘off-the-shelf’ product.

Future directions:

Few possible studies could be proposed as a continuity for this stream of novel research work performed in this thesis. However it is necessary to hypothesize a system or a bundle of tests for outcome validation as a mean for standardization of this technology. Possible methods for standardization are suggested below:

1. Decellularization device validation: a. Quantifying remaining DNA content in decellularized constructs in each future study. b. Utilize imaging techniques as confocal imaging and SEM to ensure constructs integrity and evaluate the efficiency of decellularization. 2. Implementing a Standard Operation Procedure (SOP) or a blueprint to achieve reproducible outcomes using this technology. A manual of instructions on how to reproduce this technology is found in the appendix section.

Perspective studies

• Osteoconductive or osteoinductive scaffolds should be utilized in future in vivo studies in order to improve the bone regenerative capacity.

108

• Larger animal models are required to confirm the regenerative potential of the decellularized constructs. Furthermore, the periodontal defect model should mimic as closely as possible the conditions that will be encountered clinically, and hence a plaque induced inflammatory defect model should be used, rather than a surgically created defect. • 3D decellularized constructs using a novel bi-phasic scaffold might provide simultaneous regeneration of alveolar bone and periodontal ligaments in in vivo experiments. • More primary and commercial cell lines from different tissue origins need to be tested for their potential of forming cell sheets and to be tested in vivo for their regenerative potential. • In vitro characterization and in vivo assessment of lyophilized (freeze dried) decellularized cell sheets for their biocompatibility and regenerative potential in comparison to the decellularized cell sheets investigated in the current study.

References

Adrian R Noope, Roberts AP, Yap AS, Gomez GA, Neufeld Z (2015). Modelling wound closure in an epithelial cell sheet using the cellular Potts model. Integrative biology : quantitative biosciences from nano to macro.

Ahn H, Ju YM, Takahashi H, Williams DF, Yoo JJ, Lee SJ et al. (2015). Engineered small diameter vascular grafts by combining cell sheet engineering and electrospinning technology. Acta biomaterialia 16(14-22.

Akizuki T, Oda S, Komaki M, Tsuchioka H, Kawakatsu N, Kikuchi A et al. (2005). Application of periodontal ligament cell sheet for periodontal regeneration: a pilot study in beagle dogs. Journal of periodontal research 40(3):245-251.

109

Altiere ET, Reeve CM, Sheridan PJ (1979). Lyophilized bone allografts in periodontal intraosseous defects. J Periodontol 50(10):510-519.

Armitage GC, Research S, Therapy Committee of the American Academy of P (2003). Diagnosis of periodontal diseases. J Periodontol 74(8):1237-1247.

Arnold WH, Becher S, Dannan A, Widera D, Dittmar T, Jacob M et al. (2010). Morphological characterization of periodontium-derived human stem cells. Ann Anat 192(4):215-219.

Awad SS, Rao RK, Berger DH, Albo D, Bellows CF (2009). Microbiology of infected acellular dermal matrix (AlloDerm) in patients requiring complex abdominal closure after emergency surgery. Surgical infections 10(1):79-84.

Badylak SF, Lantz GC, Coffey A, Geddes LA (1989). Small intestinal submucosa as a large diameter vascular graft in the dog. The Journal of surgical research 47(1):74-80.

Badylak SF, Tullius R, Kokini K, Shelbourne KD, Klootwyk T, Voytik SL et al. (1995). The use of xenogeneic small intestinal submucosa as a biomaterial for Achilles tendon repair in a dog model. Journal of biomedical materials research 29(8):977-985.

Badylak SF (2007). The extracellular matrix as a biologic scaffold material. Biomaterials 28(25):3587-3593.

Badylak SF, Freytes DO, Gilbert TW (2009). Extracellular matrix as a biological scaffold material: Structure and function. Acta biomaterialia 5(1):1-13.

Bai Y, Bai Y, Matsuzaka K, Hashimoto S, Fukuyama T, Wu L et al. (2011). Cementum- and periodontal ligament-like tissue formation by dental follicle cell sheets co-cultured with Hertwig's epithelial root sheath cells. Bone 48(6):1417-1426.

110

Bao J, Shi Y, Sun H, Yin X, Yang R, Li L et al. (2011). Construction of a portal implantable functional tissue-engineered liver using perfusion-decellularized matrix and hepatocytes in rats. Cell Transplant 20(5):753-766.

Barnett JD, Mellonig JT, Gray JL, Towle HJ (1989). Comparison of freeze-dried bone allograft and porous hydroxylapatite in human periodontal defects. J Periodontol 60(5):231- 237.

Bartold PM, McCulloch CA, Narayanan AS, Pitaru S (2000). Tissue engineering: a new paradigm for periodontal regeneration based on molecular and cell biology. Periodontology 2000 24(253-269.

Bartold PM, Xiao Y, Lyngstaadas SP, Paine ML, Snead ML (2006). Principles and applications of cell delivery systems for periodontal regeneration. Periodontology 2000 41(123-135.

Bashutski JD, Wang HL (2009). Periodontal and endodontic regeneration. J Endod 35(3):321-328.

Beacham DA, Amatangelo MD, Cukierman E (2007). Preparation of extracellular matrices produced by cultured and primary fibroblasts. Current protocols in cell biology / editorial board, Juan S Bonifacino [et al] Chapter 10(Unit 10 19.

Becker W, Becker BE, Berg L, Prichard J, Caffesse R, Rosenberg E (1988). New attachment after treatment with root isolation procedures: report for treated Class III and Class II furcations and vertical osseous defects. Int J Periodontics Restorative Dent 8(3):8-23.

Benders KE, van Weeren PR, Badylak SF, Saris DB, Dhert WJ, Malda J (2013). Extracellular matrix scaffolds for cartilage and bone regeneration. Trends Biotechnol 31(3):169-176.

111

Berkovitz BK (2004). Periodontal ligament: structural and clinical correlates. Dental update 31(1):46-50, 52, 54.

Bloch O, Golde P, Dohmen PM, Posner S, Konertz W, Erdbrugger W (2011). Immune response in patients receiving a bioprosthetic heart valve: lack of response with decellularized valves. Tissue Eng Part A 17(19-20):2399-2405.

Blumenthal N, Steinberg J (1990). The use of collagen membrane barriers in conjunction with combined demineralized bone-collagen gel implants in human infrabony defects. J Periodontol 61(6):319-327.

Bosshardt DD, Selvig KA (1997). Dental cementum: the dynamic tissue covering of the root. Periodontology 2000 13(41-75.

Bosshardt DD (2005). Are cementoblasts a subpopulation of osteoblasts or a unique phenotype? Journal of dental research 84(5):390-406.

Britain SK, Arx T, Schenk RK, Buser D, Nummikoski P, Cochran DL (2005). The use of guided tissue regeneration principles in endodontic surgery for induced chronic periodontic- endodontic lesions: a clinical, radiographic, and histologic evaluation. J Periodontol 76(3):450-460.

Brown AL, Brook-Allred TT, Waddell JE, White J, Werkmeister JA, Ramshaw JA et al. (2005). Bladder acellular matrix as a substrate for studying in vitro bladder smooth muscle- urothelial cell interactions. Biomaterials 26(5):529-543.

Brown BN, Valentin JE, Stewart-Akers AM, McCabe GP, Badylak SF (2009). Macrophage phenotype and remodeling outcomes in response to biologic scaffolds with and without a cellular component. Biomaterials 30(8):1482-1491.

112

Brown TD, Dalton PD, Hutmacher DW (2011). Direct writing by way of melt electrospinning. Advanced materials 23(47):5651-5657.

Bunyaratavej P, Wang HL (2001). Collagen membranes: a review. J Periodontol 72(2):215- 229.

Burk J, Erbe I, Berner D, Kacza J, Kasper C, Pfeiffer B et al. (2014). Freeze-thaw cycles enhance decellularization of large tendons. Tissue engineering Part C, Methods 20(4):276- 284.

Castrechini NM, Murthi P, Gude NM, Erwich JJ, Gronthos S, Zannettino A et al. (2010). Mesenchymal stem cells in human placental chorionic villi reside in a vascular Niche. Placenta 31(3):203-212.

Checchi L, Montevecchi M, Checchi V, Bonetti GA (2009). A modified papilla preservation technique, 22 years later. Quintessence Int 40(4):303-311.

Chen WF, Barounis D, Kalimuthu R (2010). A novel cost-saving approach to the use of acellular dermal matrix (AlloDerm) in postmastectomy breast and nipple reconstructions. Plastic and reconstructive surgery 125(2):479-481.

Cho MI, Garant PR (2000). Development and general structure of the periodontium. Periodontology 2000 24(9-27.

Cho SW, Park HJ, Ryu JH, Kim SH, Kim YH, Choi CY et al. (2005). Vascular patches tissue-engineered with autologous bone marrow-derived cells and decellularized tissue matrices. Biomaterials 26(14):1915-1924.

113

Choi DH, Suhaeri M, Hwang MP, Kim IH, Han DK, Park K (2014). Multi-lineage differentiation of human mesenchymal stromal cells on the biophysical microenvironment of cell-derived matrix. Cell Tissue Res 357(3):781-792.

Cortellini P, Pini Prato G, Baldi C, Clauser C (1990). Guided tissue regeneration with different materials. Int J Periodontics Restorative Dent 10(2):136-151.

Cortellini P, Pini Prato G, Tonetti MS (1993). Periodontal regeneration of human infrabony defects. I. Clinical measures. J Periodontol 64(4):254-260.

Cortellini P, Prato GP, Tonetti MS (1995). The modified papilla preservation technique. A new surgical approach for interproximal regenerative procedures. J Periodontol 66(4):261- 266.

Cortellini P, Pini Prato G, Tonetti MS (1996). The modified papilla preservation technique with bioresorbable barrier membranes in the treatment of intrabony defects. Case reports. Int J Periodontics Restorative Dent 16(6):546-559.

Cortellini P, Tonetti MS (2000). Focus on intrabony defects: guided tissue regeneration. Periodontology 2000 22(104-132.

Cortellini P, Tonetti MS (2005). Clinical performance of a regenerative strategy for intrabony defects: scientific evidence and clinical experience. J Periodontol 76(3):341-350.

Costa PF, Vaquette C, Zhang Q, Reis RL, Ivanovski S, Hutmacher DW (2014). Advanced tissue engineering scaffold design for regeneration of the complex hierarchical periodontal structure. J Clin Periodontol 41(3):283-294.

114

Cothren CC, Gallego K, Anderson ED, Schmidt D (2004). Chest wall reconstruction with acellular dermal matrix (AlloDerm) and a latissimus muscle flap. Plastic and reconstructive surgery 114(4):1015-1017.

Dahlin C, Linde A, Gottlow J, Nyman S (1988). Healing of bone defects by guided tissue regeneration. Plastic and reconstructive surgery 81(5):672-676.

Dan H, Vaquette C, Fisher AG, Hamlet SM, Xiao Y, Hutmacher DW et al. (2014). The influence of cellular source on periodontal regeneration using calcium phosphate coated polycaprolactone scaffold supported cell sheets. Biomaterials 35(1):113-122.

Danesh-Meyer MJ (2000). Tissue engineering in periodontics using rhBMP-2. J N Z Soc Periodontol 85):10-14.

De Sanctis M, Zucchelli G, Clauser C (1996a). Bacterial colonization of barrier material and periodontal regeneration. J Clin Periodontol 23(11):1039-1046.

De Sanctis M, Zucchelli G, Clauser C (1996b). Bacterial colonization of bioabsorbable barrier material and periodontal regeneration. J Periodontol 67(11):1193-1200.

Deliberador TM, Nagata MJ, Furlaneto FA, Melo LG, Okamoto T, Sundefeld ML et al. (2006). Autogenous bone graft with or without a calcium sulfate barrier in the treatment of Class II furcation defects: a histologic and histometric study in dogs. J Periodontol 77(5):780-789.

Demir B, Sengun D, Berberoglu A (2007). Clinical evaluation of platelet-rich plasma and bioactive glass in the treatment of intra-bony defects. J Clin Periodontol 34(8):709-715.

115

Dijkman PE, Driessen-Mol A, Frese L, Hoerstrup SP, Baaijens FP (2012). Decellularized homologous tissue-engineered heart valves as off-the-shelf alternatives to xeno- and homografts. Biomaterials 33(18):4545-4554.

Dong J, Zhang J, Gu Y (2009). [The application of decellularized matrices in the tissue engineering of vascular graft.]. Zhongguo fei ai za zhi = Chinese journal of lung cancer 12(2):152-155.

Dori F, Huszar T, Nikolidakis D, Arweiler NB, Gera I, Sculean A (2007a). Effect of platelet- rich plasma on the healing of intra-bony defects treated with a natural bone mineral and a collagen membrane. J Clin Periodontol 34(3):254-261.

Dori F, Huszar T, Nikolidakis D, Arweiler NB, Gera I, Sculean A (2007b). Effect of platelet- rich plasma on the healing of intrabony defects treated with an anorganic bovine bone mineral and expanded polytetrafluoroethylene membranes. J Periodontol 78(6):983-990.

Dori F, Huszar T, Nikolidakis D, Tihanyi D, Horvath A, Arweiler NB et al. (2008a). Effect of platelet-rich plasma on the healing of intrabony defects treated with Beta tricalcium phosphate and expanded polytetrafluoroethylene membranes. J Periodontol 79(4):660-669.

Dori F, Nikolidakis D, Huszar T, Arweiler NB, Gera I, Sculean A (2008b). Effect of platelet- rich plasma on the healing of intrabony defects treated with an enamel matrix protein derivative and a natural bone mineral. J Clin Periodontol 35(1):44-50.

Dragoo MR, Sullivan HC (1973). A clinical and histological evaluation of autogenous iliac bone grafts in humans. II. External root resorption. J Periodontol 44(10):614-625.

Elder BD, Eleswarapu SV, Athanasiou KA (2009). Extraction techniques for the decellularization of tissue engineered articular cartilage constructs. Biomaterials 30(22):3749-3756.

116

Esposito M, Grusovin MG, Papanikolaou N, Coulthard P, Worthington HV (2009). Enamel matrix derivative (Emdogain(R)) for periodontal tissue regeneration in intrabony defects. Cochrane Database Syst Rev 4):CD003875.

Fang NT, Xie SZ, Wang SM, Gao HY, Wu CG, Pan LF (2007). Construction of tissue- engineered heart valves by using decellularized scaffolds and endothelial progenitor cells. Chin Med J (Engl) 120(8):696-702.

Farag A, Vaquette C, Theodoropoulos C, Hamlet SM, Hutmacher DW, Ivanovski S (2014). Decellularized periodontal ligament cell sheets with recellularization potential. Journal of dental research 93(12):1313-1319.

Faulk DM, Carruthers CA, Warner HJ, Kramer CR, Reing JE, Zhang L et al. (2014). The effect of detergents on the basement membrane complex of a biologic scaffold material. Acta biomaterialia 10(1):183-193.

Flores MG, Hasegawa M, Yamato M, Takagi R, Okano T, Ishikawa I (2008a). Cementum- periodontal ligament complex regeneration using the cell sheet technique. Journal of periodontal research 43(3):364-371.

Flores MG, Yashiro R, Washio K, Yamato M, Okano T, Ishikawa I (2008b). Periodontal ligament cell sheet promotes periodontal regeneration in athymic rats. J Clin Periodontol 35(12):1066-1072.

Flynn L, Semple JL, Woodhouse KA (2006). Decellularized placental matrices for adipose tissue engineering. Journal of biomedical materials research Part A 79(2):359-369.

117

Froum SJ, Weinberg MA, Rosenberg E, Tarnow D (2001). A comparative study utilizing open flap debridement with and without enamel matrix derivative in the treatment of periodontal intrabony defects: a 12-month re-entry study. J Periodontol 72(1):25-34.

Gao H, Li B, Zhao L, Jin Y (2015). Influence of nanotopography on periodontal ligament stem cell functions and cell sheet based periodontal regeneration. Int J Nanomedicine 10(4009-4027.

Gawlitta D, Benders KE, Visser J, van der Sar AS, Kempen DH, Theyse LF et al. (2015). Decellularized cartilage-derived matrix as substrate for endochondral bone regeneration. Tissue Eng Part A 21(3-4):694-703.

Gestrelius S, Andersson C, Johansson AC, Persson E, Brodin A, Rydhag L et al. (1997). Formulation of enamel matrix derivative for surface coating. Kinetics and cell colonization. J Clin Periodontol 24(9 Pt 2):678-684.

Giannobile WV, Somerman MJ (2003). Growth and amelogenin-like factors in periodontal wound healing. A systematic review. Ann Periodontol 8(1):193-204.

Gilbert TW, Sellaro TL, Badylak SF (2006). Decellularization of tissues and organs. Biomaterials 27(19):3675-3683.

Gottlow J, Nyman S, Lindhe J, Karring T, Wennstrom J (1986). New attachment formation in the human periodontium by guided tissue regeneration. Case reports. J Clin Periodontol 13(6):604-616.

Gottlow J, Karring T, Nyman S (1990). Guided tissue regeneration following treatment of recession-type defects in the monkey. J Periodontol 61(11):680-685.

118

Grandi C, Baiguera S, Martorina F, Lora S, Amista P, Dalzoppo D et al. (2011). Decellularized bovine reinforced vessels for small-diameter tissue-engineered vascular grafts. Int J Mol Med 28(3):315-325.

Groeger S, Doman E, Chakraborty T, Meyle J (2010). Effects of Porphyromonas gingivalis infection on human gingival epithelial barrier function in vitro. European journal of oral sciences 118(6):582-589.

Haag J, Baiguera S, Jungebluth P, Barale D, Del Gaudio C, Castiglione F et al. (2012). Biomechanical and angiogenic properties of tissue-engineered rat trachea using genipin cross-linked decellularized tissue. Biomaterials 33(3):780-789.

Hagewald S, Pischon N, Jawor P, Bernimoulin JP, Zimmermann B (2004). Effects of enamel matrix derivative on proliferation and differentiation of primary osteoblasts. Oral surgery, oral medicine, oral pathology, oral radiology, and endodontics 98(2):243-249.

He J, Jiang J, Safavi KE, Spangberg LS, Zhu Q (2004). Direct contact between enamel matrix derivative (EMD) and osteoblasts is not required for EMD-induced cell proliferation. Oral surgery, oral medicine, oral pathology, oral radiology, and endodontics 98(3):370-375.

Heard RH, Mellonig JT, Brunsvold MA, Lasho DJ, Meffert RM, Cochran DL (2000). Clinical evaluation of wound healing following multiple exposures to enamel matrix protein derivative in the treatment of intrabony periodontal defects. J Periodontol 71(11):1715-1721.

Hesami P, Holzapfel BM, Taubenberger A, Roudier M, Fazli L, Sieh S et al. (2014). A humanized tissue-engineered in vivo model to dissect interactions between human prostate cancer cells and human bone. Clin Exp Metastasis.

Hiatt WH, Schallhorn RG (1973). Intraoral transplants of cancellous bone and marrow in periodontal lesions. J Periodontol 44(4):194-208.

119

Hirabaru M, Kuroki T, Adachi T, Kitasato A, Ono S, Tanaka T et al. (2015). A Method for Performing Islet Transplantation Using Tissue-Engineered Sheets of Islets and Mesenchymal Stem Cells. Tissue engineering Part C, Methods.

Hoshiba T, Lu H, Kawazoe N, Chen G (2010). Decellularized matrices for tissue engineering. Expert opinion on biological therapy 10(12):1717-1728.

Houser BE, Mellonig JT, Brunsvold MA, Cochran DL, Meffert RM, Alder ME (2001). Clinical evaluation of anorganic bovine bone xenograft with a bioabsorbable collagen barrier in the treatment of molar furcation defects. Int J Periodontics Restorative Dent 21(2):161- 169.

Ishikawa I, Iwata T, Washio K, Okano T, Nagasawa T, Iwasaki K et al. (2009). Cell sheet engineering and other novel cell-based approaches to periodontal regeneration. Periodontology 2000 51(220-238.

Ivanovski S, Li H, Haase HR, Bartold PM (2001). Expression of bone associated macromolecules by gingival and periodontal ligament fibroblasts. Journal of periodontal research 36(3):131-141.

Ivanovski S, Gronthos S, Shi S, Bartold PM (2006). Stem cells in the periodontal ligament. Oral Dis 12(4):358-363.

Ivanovski S (2009). Periodontal regeneration. Aust Dent J 54 Suppl 1(S118-128.

Izumi Y, Aoki A, Yamada Y, Kobayashi H, Iwata T, Akizuki T et al. (2011). Current and future periodontal tissue engineering. Periodontology 2000 56(1):166-187.

120

Jin QM, Anusaksathien O, Webb SA, Rutherford RB, Giannobile WV (2003). Gene therapy of bone morphogenetic protein for periodontal tissue engineering. J Periodontol 74(2):202- 213.

Kadowaki T, Yamamoto K (2003). Suppression of virulence of Porphyromonas gingivalis by potent inhibitors specific for gingipains. Current protein & peptide science 4(6):451-458.

Kaigler D, Cirelli JA, Giannobile WV (2006). Growth factor delivery for oral and periodontal tissue engineering. Expert Opin Drug Deliv 3(5):647-662.

Kao RT, Murakami S, Beirne OR (2009). The use of biologic mediators and tissue engineering in dentistry. Periodontology 2000 50(127-153.

Kawaguchi H, Hayashi H, Mizuno N, Fujita T, Hasegawa N, Shiba H et al. (2005). [Cell transplantation for periodontal diseases. A novel periodontal tissue regenerative therapy using bone marrow mesenchymal stem cells]. Clin Calcium 15(7):99-104.

Kim CS, Um YJ, Chai JK, Cho KS, Moon IS, Choi SH et al. (2011). A canine model for histometric evaluation of periodontal regeneration. Periodontology 2000 56(1):209-226.

Kish KJ, Buinewicz BR, Morris JB (2005). Acellular dermal matrix (AlloDerm): new material in the repair of stoma site hernias. The American surgeon 71(12):1047-1050.

Kitamura M, Nakashima K, Kowashi Y, Fujii T, Shimauchi H, Sasano T et al. (2008). Periodontal tissue regeneration using fibroblast growth factor-2: randomized controlled phase II clinical trial. PloS one 3(7):e2611.

Ko YM, Choi DY, Jung SC, Kim BH (2015). Characteristics of Plasma Treated Electrospun Polycaprolactone (PCL) Nanofiber Scaffold for Bone Tissue Engineering. J Nanosci Nanotechnol 15(1):192-195.

121

Kobayashi T, Kan K, Nishida K, Yamato M, Okano T (2013). Corneal regeneration by transplantation of corneal epithelial cell sheets fabricated with automated cell culture system in rabbit model. Biomaterials 34(36):9010-9017.

Korn JM, Connolly MM, Walton RL (2009). Single-stage sacral coccygectomy and repair using human acellular dermal matrix (AlloDerm) with bilateral gluteus maximus flaps for hernia prophylaxis. Hernia : the journal of hernias and abdominal wall surgery 13(3):329- 332.

Kropp BP, Eppley BL, Prevel CD, Rippy MK, Harruff RC, Badylak SF et al. (1995). Experimental assessment of small intestinal submucosa as a bladder wall substitute. Urology 46(3):396-400.

Lam CXF, Teoh SH, Hutmacher DW (2007). Comparison of the degradation of polycaprolactone and polycaprolactone–(β-tricalcium phosphate) scaffolds in alkaline medium. Polymer International 56(6):718-728.

Langer R, Vacanti JP (1993). Tissue engineering. Science 260(5110):920-926.

Lin G, Liu G, Banie L, Wang G, Ning H, Lue TF et al. (2011). Tissue distribution of mesenchymal stem cell marker Stro-1. Stem Cells Dev 20(10):1747-1752.

Lin NH, Gronthos S, Bartold PM (2008). Stem cells and periodontal regeneration. Aust Dent J 53(2):108-121.

Lin NH, Gronthos S, Mark Bartold P (2009). Stem cells and future periodontal regeneration. Periodontology 2000 51(239-251.

122

Lindhe J, Pontoriero R, Berglundh T, Araujo M (1995). The effect of flap management and bioresorbable occlusive devices in GTR treatment of degree III furcation defects. An experimental study in dogs. J Clin Periodontol 22(4):276-283.

Lu H, Hoshiba T, Kawazoe N, Chen G (2012). Comparison of decellularization techniques for preparation of extracellular matrix scaffolds derived from three-dimensional cell culture. Journal of biomedical materials research Part A 100(9):2507-2516.

Ma J, Smietana MJ, Kostrominova TY, Wojtys EM, Larkin LM, Arruda EM (2012). Three- dimensional engineered bone-ligament-bone constructs for anterior cruciate ligament replacement. Tissue Eng Part A 18(1-2):103-116.

Maeda H, Tomokiyo A, Fujii S, Wada N, Akamine A (2011). Promise of periodontal ligament stem cells in regeneration of periodontium. Stem Cell Res Ther 2(4):33.

Manor A (2000). Periodontal regeneration with enamel matrix derivative--case reports. Journal of the International Academy of Periodontology 2(2):44-48.

Mao JJ, Giannobile WV, Helms JA, Hollister SJ, Krebsbach PH, Longaker MT et al. (2006). Craniofacial tissue engineering by stem cells. Journal of dental research 85(11):966-979.

Markou N, Pepelassi E, Vavouraki H, Stamatakis HC, Nikolopoulos G, Vrotsos I et al. (2009). Treatment of periodontal endosseous defects with platelet-rich plasma alone or in combination with demineralized freeze-dried bone allograft: a comparative clinical trial. J Periodontol 80(12):1911-1919.

Marx RE (2004). Platelet-rich plasma: evidence to support its use. Journal of oral and maxillofacial surgery : official journal of the American Association of Oral and Maxillofacial Surgeons 62(4):489-496.

123

McClain PK, Schallhorn RG (1993). Long-term assessment of combined osseous composite grafting, root conditioning, and guided tissue regeneration. Int J Periodontics Restorative Dent 13(1):9-27.

McClain PK, Schallhorn RG (2000). Focus on furcation defects--guided tissue regeneration in combination with bone grafting. Periodontology 2000 22(190-212.

McGuire MK, Kao RT, Nevins M, Lynch SE (2006). rhPDGF-BB promotes healing of periodontal defects: 24-month clinical and radiographic observations. Int J Periodontics Restorative Dent 26(3):223-231.

Mellonig JT, Bowers GM, Bright RW, Lawrence JJ (1976). Clinical evaluation of freeze- dried bone allografts in periodontal osseous defects. J Periodontol 47(3):125-131.

Mellonig JT (1984). Decalcified freeze-dried bone allograft as an implant material in human periodontal defects. Int J Periodontics Restorative Dent 4(6):40-55.

Mellonig JT, Bowers GM (1990). Regenerating bone in clinical periodontics. J Am Dent Assoc 121(4):497-502.

Mellonig JT (1991). Freeze-dried bone allografts in periodontal reconstructive surgery. Dent Clin North Am 35(3):505-520.

Mellonig JT (1999). Enamel matrix derivative for periodontal reconstructive surgery: technique and clinical and histologic case report. Int J Periodontics Restorative Dent 19(1):8- 19.

Miura M, Gronthos S, Zhao M, Lu B, Fisher LW, Robey PG et al. (2003). SHED: stem cells from human exfoliated deciduous teeth. Proc Natl Acad Sci U S A 100(10):5807-5812.

124

Murakami S, Takayama S, Kitamura M, Shimabukuro Y, Yanagi K, Ikezawa K et al. (2003a). Recombinant human basic fibroblast growth factor (bFGF) stimulates periodontal regeneration in class II furcation defects created in beagle dogs. Journal of periodontal research 38(1):97-103.

Murakami S (2011). Periodontal tissue regeneration by signaling molecule(s): what role does basic fibroblast growth factor (FGF-2) have in periodontal therapy? Periodontology 2000 56(1):188-208.

Murakami Y, Kojima T, Nagasawa T, Kobayashi H, Ishikawa I (2003b). Novel isolation of alkaline phosphatase-positive subpopulation from periodontal ligament fibroblasts. J Periodontol 74(6):780-786.

Murase Y, Narita Y, Kagami H, Miyamoto K, Ueda Y, Ueda M et al. (2006). Evaluation of compliance and stiffness of decellularized tissues as scaffolds for tissue-engineered small caliber vascular grafts using intravascular ultrasound. ASAIO J 52(4):450-455.

Murphy KG (1995). Postoperative healing complications associated with Gore-Tex Periodontal Material. Part II. Effect of complications on regeneration. Int J Periodontics Restorative Dent 15(6):548-561.

Nakajima R, Kobayashi T, Moriya N, Mizutani M, Kan K, Nozaki T et al. (2012). A novel closed cell culture device for fabrication of corneal epithelial cell sheets. J Tissue Eng Regen Med.

Nakajima R, Kobayashi T, Kikuchi T, Kitano Y, Watanabe H, Mizutani M et al. (2015). Fabrication of transplantable corneal epithelial and oral mucosal epithelial cell sheets using a novel temperature-responsive closed culture device. J Tissue Eng Regen Med 9(5):637-640.

125

Nakayama KH, Lee CC, Batchelder CA, Tarantal AF (2013). Tissue specificity of decellularized rhesus monkey kidney and lung scaffolds. PloS one 8(5):e64134.

Nanci A, Bosshardt DD (2006a). Structure of periodontal tissues in health and disease. Periodontology 2000 40(11-28.

Nanci A, Bosshardt DD (2006b). Structure of periodontal tissues in health and disease. Periodontology 2000 40(1):11-28.

Needleman IG, Worthington HV, Giedrys-Leeper E, Tucker RJ (2006). Guided tissue regeneration for periodontal infra-bony defects. Cochrane Database Syst Rev 2):CD001724.

Nevins M, Giannobile WV, McGuire MK, Kao RT, Mellonig JT, Hinrichs JE et al. (2005). Platelet-derived growth factor stimulates bone fill and rate of attachment level gain: results of a large multicenter randomized controlled trial. J Periodontol 76(12):2205-2215.

Nevins M, Nevins ML, Kim SW, Schupbach P, Kim DM (2011). The use of mucograft collagen matrix to augment the zone of keratinized tissue around teeth: a pilot study. Int J Periodontics Restorative Dent 31(4):367-373.

Ning H, Lin G, Lue TF, Lin CS (2011). Mesenchymal stem cell marker Stro-1 is a 75 kd endothelial antigen. Biochemical and biophysical research communications 413(2):353-357.

NM OB-S, Veith PD, Dashper SG, Reynolds EC (2003). Porphyromonas gingivalis gingipains: the molecular teeth of a microbial vampire. Current protein & peptide science 4(6):409-426.

Nonaka PN, Campillo N, Uriarte JJ, Garreta E, Melo E, de Oliveira LV et al. (2014). Effects of freezing/thawing on the mechanical properties of decellularized lungs. Journal of biomedical materials research Part A 102(2):413-419.

126

Nowzari H, Matian F, Slots J (1995). Periodontal pathogens on polytetrafluoroethylene membrane for guided tissue regeneration inhibit healing. J Clin Periodontol 22(6):469-474.

Nyman S, Lindhe J, Karring T, Rylander H (1982). New attachment following surgical treatment of human periodontal disease. J Clin Periodontol 9(4):290-296.

Nyman S, Gottlow J, Lindhe J, Karring T, Wennstrom J (1987). New attachment formation by guided tissue regeneration. Journal of periodontal research 22(3):252-254.

Okuda K, Tai H, Tanabe K, Suzuki H, Sato T, Kawase T et al. (2005). Platelet-rich plasma combined with a porous hydroxyapatite graft for the treatment of intrabony periodontal defects in humans: a comparative controlled clinical study. J Periodontol 76(6):890-898.

Ouyang XY, Qiao J (2006). Effect of platelet-rich plasma in the treatment of periodontal intrabony defects in humans. Chin Med J (Engl) 119(18):1511-1521.

Papadimitropoulos A, Scotti C, Bourgine P, Scherberich A, Martin I (2015). Engineered decellularized matrices to instruct bone regeneration processes. Bone 70(66-72.

Patel J, Shafiee A, Wang W, Fisk NM, Khosrotehrani K (2014). Novel isolation strategy to deliver pure fetal-origin and maternal-origin mesenchymal stem cell (MSC) populations from human term placenta. Placenta 35(11):969-971.

Pearson GE, Rosen S, Deporter DA (1981). Preliminary observations on the usefulness of a decalcified, freeze-dried cancellous bone allograft material in periodontal surgery. J Periodontol 52(2):55-59.

127

Piattelli A, Scarano A, Corigliano M, Piattelli M (1996). Comparison of bone regeneration with the use of mineralized and demineralized freeze-dried bone allografts: a histological and histochemical study in man. Biomaterials 17(11):1127-1131.

Pitt CG, Gratzl MM, Kimmel GL, Surles J, Schindler A (1981). Aliphatic polyesters II. The degradation of poly (DL-lactide), poly (epsilon-caprolactone), and their copolymers in vivo. Biomaterials 2(4):215-220.

Polimeni G, Xiropaidis AV, Wikesjo UM (2006). Biology and principles of periodontal wound healing/regeneration. Periodontology 2000 41(30-47.

Pollanen MT, Salonen JI, Grenier D, Uitto VJ (2000). Epithelial cell response to challenge of bacterial lipoteichoic acids and lipopolysaccharides in vitro. Journal of medical microbiology 49(3):245-252.

Potempa J, Banbula A, Travis J (2000). Role of bacterial proteinases in matrix destruction and modulation of host responses. Periodontology 2000 24(153-192.

Quint C, Kondo Y, Manson RJ, Lawson JH, Dardik A, Niklason LE (2011). Decellularized tissue-engineered blood vessel as an arterial conduit. Proc Natl Acad Sci U S A 108(22):9214-9219.

Rahaman MN, Mao JJ (2005). Stem cell-based composite tissue constructs for regenerative medicine. Biotechnol Bioeng 91(3):261-284.

Rasperini G, Silvestri M, Schenk RK, Nevins ML (2000). Clinical and histologic evaluation of human gingival recession treated with a subepithelial connective tissue graft and enamel matrix derivative (Emdogain): a case report. Int J Periodontics Restorative Dent 20(3):269- 275.

128

Rathe F, Junker R, Chesnutt BM, Jansen JA (2009). The effect of enamel matrix derivative (Emdogain) on bone formation: a systematic review. Tissue engineering Part B, Reviews 15(3):215-224.

Reichert da Silva Assuncao L, Colenci R, Ferreira do-Amaral CC, Sonoda CK, Mogami Bomfim SR, Okamoto R et al. (2011). Periodontal tissue engineering after tooth replantation. J Periodontol 82(5):758-766.

Reichert JC, Quent VM, Burke LJ, Stansfield SH, Clements JA, Hutmacher DW (2010). Mineralized human primary osteoblast matrices as a model system to analyse interactions of prostate cancer cells with the bone microenvironment. Biomaterials 31(31):7928-7936.

Remya KR, Joseph J, Mani S, John A, Varma HK, Ramesh P (2013). Nanohydroxyapatite incorporated electrospun polycaprolactone/polycaprolactone-polyethyleneglycol- polycaprolactone blend scaffold for bone tissue engineering applications. J Biomed Nanotechnol 9(9):1483-1494.

Reynolds MA, Aichelmann-Reidy ME, Branch-Mays GL, Gunsolley JC (2003). The efficacy of bone replacement grafts in the treatment of periodontal osseous defects. A systematic review. Ann Periodontol 8(1):227-265.

Rodrigues SV, Acharya AB, Thakur SL (2012). Platelet-rich plasma. A review. The New York state dental journal 78(1):26-30.

Rosen PS, Reynolds MA (2002). A retrospective case series comparing the use of demineralized freeze-dried bone allograft and freeze-dried bone allograft combined with enamel matrix derivative for the treatment of advanced osseous lesions. J Periodontol 73(8):942-949.

129

Rotundo R, Pini-Prato G (2012). Use of a new collagen matrix (mucograft) for the treatment of multiple gingival recessions: case reports. Int J Periodontics Restorative Dent 32(4):413- 419.

Sadr N, Pippenger BE, Scherberich A, Wendt D, Mantero S, Martin I et al. (2012). Enhancing the biological performance of synthetic polymeric materials by decoration with engineered, decellularized extracellular matrix. Biomaterials 33(20):5085-5093.

Salzberg CA, Ashikari AY, Koch RM, Chabner-Thompson E (2011). An 8-year experience of direct-to-implant immediate breast reconstruction using human acellular dermal matrix (AlloDerm). Plastic and reconstructive surgery 127(2):514-524.

Sanders JJ, Sepe WW, Bowers GM, Koch RW, Williams JE, Lekas JS et al. (1983). Clinical evaluation of freeze-dried bone allografts in periodontal osseous defects. Part III. Composite freeze-dried bone allografts with and without autogenous bone grafts. J Periodontol 54(1):1- 8.

Sanz M, Giovannoli JL (2000). Focus on furcation defects: guided tissue regeneration. Periodontology 2000 22(169-189.

Sanz M, Lorenzo R, Aranda JJ, Martin C, Orsini M (2009). Clinical evaluation of a new collagen matrix (Mucograft prototype) to enhance the width of keratinized tissue in patients with fixed prosthetic restorations: a randomized prospective clinical trial. J Clin Periodontol 36(10):868-876.

Sawyer AA, Song SJ, Susanto E, Chuan P, Lam CX, Woodruff MA et al. (2009). The stimulation of healing within a rat calvarial defect by mPCL-TCP/collagen scaffolds loaded with rhBMP-2. Biomaterials 30(13):2479-2488.

130

Schallhorn RG (1968). The use of autogenous hip marrow biopsy implants for bony crater defects. J Periodontol 39(3):145-147.

Schappacher M, Le Hellaye M, Bareille R, Durrieu MC, Guillaume SM (2010). Comparative in vitro cytotoxicity toward human osteoprogenitor cells of polycaprolactones synthesized from various metallic initiators. Macromol Biosci 10(1):60-67.

Scheyer ET, Velasquez-Plata D, Brunsvold MA, Lasho DJ, Mellonig JT (2002). A clinical comparison of a bovine-derived xenograft used alone and in combination with enamel matrix derivative for the treatment of periodontal osseous defects in humans. J Periodontol 73(4):423-432.

Schlueter SR, Carnes DL, Cochran DL (2007). In vitro effects of enamel matrix derivative on microvascular cells. J Periodontol 78(1):141-151.

Schmidt D, Stock UA, Hoerstrup SP (2007). Tissue engineering of heart valves using decellularized xenogeneic or polymeric starter matrices. Philosophical transactions of the Royal Society of London Series B, Biological sciences 362(1484):1505-1512.

Schwartz Z, Mellonig JT, Carnes DL, Jr., de la Fontaine J, Cochran DL, Dean DD et al. (1996). Ability of commercial demineralized freeze-dried bone allograft to induce new bone formation. J Periodontol 67(9):918-926.

Schwartz Z, Somers A, Mellonig JT, Carnes DL, Jr., Wozney JM, Dean DD et al. (1998). Addition of human recombinant bone morphogenetic protein-2 to inactive commercial human demineralized freeze-dried bone allograft makes an effective composite bone inductive implant material. J Periodontol 69(12):1337-1345.

131

Sculean A, Chiantella GC, Windisch P, Donos N (2000). Clinical and histologic evaluation of human intrabony defects treated with an enamel matrix protein derivative (Emdogain). Int J Periodontics Restorative Dent 20(4):374-381.

Sculean A, Chiantella GC, Windisch P, Gera I, Reich E (2002). Clinical evaluation of an enamel matrix protein derivative (Emdogain) combined with a bovine-derived xenograft (Bio-Oss) for the treatment of intrabony periodontal defects in humans. Int J Periodontics Restorative Dent 22(3):259-267.

Sculean A, Berakdar M, Chiantella GC, Donos N, Arweiler NB, Brecx M (2003). Healing of intrabony defects following treatment with a bovine-derived xenograft and collagen membrane. A controlled clinical study. J Clin Periodontol 30(1):73-80.

Sculean A, Stavropoulos A, Windisch P, Keglevich T, Karring T, Gera I (2004a). Healing of human intrabony defects following regenerative periodontal therapy with a bovine-derived xenograft and guided tissue regeneration. Clin Oral Investig 8(2):70-74.

Sculean A, Windisch P, Chiantella GC (2004b). Human histologic evaluation of an intrabony defect treated with enamel matrix derivative, xenograft, and GTR. Int J Periodontics Restorative Dent 24(4):326-333.

Sculean A, Windisch P, Keglevich T, Gera I (2005). Clinical and histologic evaluation of an enamel matrix protein derivative combined with a bioactive glass for the treatment of intrabony periodontal defects in humans. Int J Periodontics Restorative Dent 25(2):139-147.

Sculean A, Pietruska M, Arweiler NB, Auschill TM, Nemcovsky C (2007a). Four-year results of a prospective-controlled clinical study evaluating healing of intra-bony defects following treatment with an enamel matrix protein derivative alone or combined with a bioactive glass. J Clin Periodontol 34(6):507-513.

132

Sculean A, Schwarz F, Becker J, Brecx M (2007b). The application of an enamel matrix protein derivative (Emdogain) in regenerative periodontal therapy: a review. Medical principles and practice : international journal of the Kuwait University, Health Science Centre 16(3):167-180.

Selvig KA, Kersten BG, Chamberlain AD, Wikesjo UM, Nilveus RE (1992). Regenerative surgery of intrabony periodontal defects using ePTFE barrier membranes: scanning electron microscopic evaluation of retrieved membranes versus clinical healing. J Periodontol 63(12):974-978.

Seo BM, Miura M, Gronthos S, Bartold PM, Batouli S, Brahim J et al. (2004). Investigation of multipotent postnatal stem cells from human periodontal ligament. Lancet 364(9429):149- 155.

Shao Z, Zhang X, Pi Y, Yin L, Li L, Chen H et al. (2015). Surface modification on polycaprolactone electrospun mesh and human decalcified bone scaffold with synovium- derived mesenchymal stem cells-affinity peptide for tissue engineering. Journal of biomedical materials research Part A 103(1):318-329.

Shimizu K, Ito A, Arinobe M, Murase Y, Iwata Y, Narita Y et al. (2007). Effective cell- seeding technique using magnetite nanoparticles and magnetic force onto decellularized blood vessels for vascular tissue engineering. Journal of bioscience and bioengineering 103(5):472-478.

Sigurdsson TJ, Lee MB, Kubota K, Turek TJ, Wozney JM, Wikesjo UM (1995). Periodontal repair in dogs: recombinant human bone morphogenetic protein-2 significantly enhances periodontal regeneration. J Periodontol 66(2):131-138.

Stavropoulos A, Karring T (2004). Long-term stability of periodontal conditions achieved following guided tissue regeneration with bioresorbable membranes: case series results after 6-7 years. J Clin Periodontol 31(11):939-944.

133

Steinberg AD, LeBreton G, Willey R, Mukherjee S, Lipowski J (1986). Extravascular clot formation and platelet activation on variously treated root surfaces. J Periodontol 57(8):516- 522.

Sutherland AJ, Beck EC, Dennis SC, Converse GL, Hopkins RA, Berkland CJ et al. (2015). Decellularized cartilage may be a chondroinductive material for osteochondral tissue engineering. PloS one 10(5):e0121966.

Syed O, Walters NJ, Day RM, Kim HW, Knowles JC (2014). Evaluation of decellularization protocols for production of tubular small intestine submucosa scaffolds for use in oesophageal tissue engineering. Acta biomaterialia.

Syedain ZH, Bradee AR, Kren S, Taylor DA, Tranquillo RT (2013). Decellularized tissue- engineered heart valve leaflets with recellularization potential. Tissue Eng Part A 19(5- 6):759-769.

Takahashi H, Shigeta T, Umeda M, Komori T (2011). A new in vitro invasion model for oral cancer using an acellular allogenic dermal matrix (Alloderm): the relationship among in vitro invasion activity, in vivo invasion and metastasis. The Kobe journal of medical sciences 57(4):E128-136.

Takeda YS, Xu Q (2014). Fabrication of 2D and 3D constructs from reconstituted decellularized tissue extracellular matrices. J Biomed Nanotechnol 10(12):3631-3637.

Tambralli A, Blakeney B, Anderson J, Kushwaha M, Andukuri A, Dean D et al. (2009). A hybrid biomimetic scaffold composed of electrospun polycaprolactone nanofibers and self- assembled peptide amphiphile nanofibers. Biofabrication 1(2):025001.

134

Tang C, Xu Y, Jin C, Min BH, Li Z, Pei X et al. (2013). Feasibility of autologous bone marrow mesenchymal stem cell-derived extracellular matrix scaffold for cartilage tissue engineering. Artif Organs 37(12):E179-190.

Tapias LF, Ott HC (2014). Decellularized scaffolds as a platform for bioengineered organs. Curr Opin Organ Transplant 19(2):145-152.

Taubenberger AV, Quent VM, Thibaudeau L, Clements JA, Hutmacher DW (2013). Delineating breast cancer cell interactions with engineered bone microenvironments. J Bone Miner Res 28(6):1399-1411.

Thakkar S, Ghebes CA, Ahmed M, Kelder C, van Blitterswijk CA, Saris D et al. (2013). Mesenchymal stromal cell-derived extracellular matrix influences gene expression of chondrocytes. Biofabrication 5(2):025003.

Thibaudeau L, Quent VM, Holzapfel BM, Taubenberger AV, Straub M, Hutmacher DW (2014a). Mimicking breast cancer-induced bone metastasis in vivo: current transplantation models and advanced humanized strategies. Cancer Metastasis Rev 33(2-3):721-735.

Thibaudeau L, Taubenberger AV, Holzapfel BM, Quent VM, Fuehrmann T, Hesami P et al. (2014b). A tissue-engineered humanized xenograft model of human breast cancer metastasis to bone. Disease models & mechanisms 7(2):299-309.

Tomar GB, Srivastava RK, Gupta N, Barhanpurkar AP, Pote ST, Jhaveri HM et al. (2010). Human gingiva-derived mesenchymal stem cells are superior to bone marrow-derived mesenchymal stem cells for cell therapy in regenerative medicine. Biochemical and biophysical research communications 393(3):377-383.

135

Trombelli L, Farina R (2008). Clinical outcomes with bioactive agents alone or in combination with grafting or guided tissue regeneration. J Clin Periodontol 35(8 Suppl):117- 135.

Trubiani O, Orsini G, Zini N, Di Iorio D, Piccirilli M, Piattelli A et al. (2008). Regenerative potential of human periodontal ligament derived stem cells on three-dimensional biomaterials: a morphological report. Journal of biomedical materials research Part A 87(4):986-993.

Tsumanuma Y, Iwata T, Washio K, Yoshida T, Yamada A, Takagi R et al. (2011). Comparison of different tissue-derived stem cell sheets for periodontal regeneration in a canine 1-wall defect model. Biomaterials 32(25):5819-5825.

Tu Q, Zhang Y, Li Y, Chen H (2007). [Fabrication and evaluation of a new decellularized tissue engineered blood vessel scaffold]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 24(2):379-384.

Urist MR (1965). Bone: formation by autoinduction. Science 150(3698):893-899.

Vaquette C, Cooper-White JJ (2011). Increasing electrospun scaffold pore size with tailored collectors for improved cell penetration. Acta biomaterialia 7(6):2544-2557.

Vaquette C, Fan W, Xiao Y, Hamlet S, Hutmacher DW, Ivanovski S (2012). A biphasic scaffold design combined with cell sheet technology for simultaneous regeneration of alveolar bone/periodontal ligament complex. Biomaterials 33(22):5560-5573.

Vaquette C, Ivanovski S, Hamlet SM, Hutmacher DW (2013). Effect of culture conditions and calcium phosphate coating on ectopic bone formation. Biomaterials 34(22):5538-5551.

136

Wainwright DJ (1995). Use of an acellular allograft dermal matrix (AlloDerm) in the management of full-thickness burns. Burns : journal of the International Society for Burn Injuries 21(4):243-248.

Weber B, Dijkman PE, Scherman J, Sanders B, Emmert MY, Grunenfelder J et al. (2013). Off-the-shelf human decellularized tissue-engineered heart valves in a non-human primate model. Biomaterials 34(30):7269-7280.

Weymann A, Patil NP, Sabashnikov A, Korkmaz S, Li S, Soos P et al. (2015). Perfusion- Decellularization of Porcine Lung and Trachea for Respiratory Bioengineering. Artif Organs.

White JK, Agnihotri AK, Titus JS, Torchiana DF (2005). A stentless trileaflet valve from a sheet of decellularized porcine small intestinal submucosa. The Annals of thoracic surgery 80(2):704-707.

Widera D, Grimm WD, Moebius JM, Mikenberg I, Piechaczek C, Gassmann G et al. (2007). Highly efficient neural differentiation of human somatic stem cells, isolated by minimally invasive periodontal surgery. Stem Cells Dev 16(3):447-460.

Woodruff MA, Hutmacher DW (2010). The return of a forgotten polymer—Polycaprolactone in the 21st century. Progress in Polymer Science 35(10):1217-1256.

Wu Q, Bao J, Zhou YJ, Wang YJ, Du ZG, Shi YJ et al. (2015). Optimizing perfusion- decellularization methods of porcine livers for clinical-scale whole-organ bioengineering. Biomed Res Int 2015(785474.

Wulf GG, Viereck V, Hemmerlein B, Haase D, Vehmeyer K, Pukrop T et al. (2004). Mesengenic progenitor cells derived from human placenta. Tissue Eng 10(7-8):1136-1147.

137

Xing Q, Yates K, Tahtinen M, Shearier E, Qian Z, Zhao F (2015). Decellularization of fibroblast cell sheets for natural extracellular matrix scaffold preparation. Tissue engineering Part C, Methods 21(1):77-87.

Xiong Y, Chan WY, Chua AW, Feng J, Gopal P, Ong YS et al. (2013). Decellularized porcine saphenous artery for small-diameter tissue-engineered conduit graft. Artif Organs 37(6):E74-87.

Yazdani SK, Watts B, Machingal M, Jarajapu YP, Van Dyke ME, Christ GJ (2009). Smooth muscle cell seeding of decellularized scaffolds: the importance of bioreactor preconditioning to development of a more native architecture for tissue-engineered blood vessels. Tissue Eng Part A 15(4):827-840.

Yilmaz S, Cakar G, Kuru B, Dirikan S, Yildirim B (2009). Platelet-rich plasma in combination with bovine derived xenograft in the treatment of deep intrabony periodontal defects: a report of 20 consecutively treated patients. Platelets 20(6):432-440.

Yin Z, Chen X, Zhu T, Hu JJ, Song HX, Shen WL et al. (2013). The effect of decellularized matrices on human tendon stem/progenitor cell differentiation and tendon repair. Acta biomaterialia 9(12):9317-9329.

Yoon H, Kim G (2011). A three-dimensional polycaprolactone scaffold combined with a drug delivery system consisting of electrospun nanofibers. J Pharm Sci 100(2):424-430.

Yukna RA, Mellonig JT (2000). Histologic evaluation of periodontal healing in humans following regenerative therapy with enamel matrix derivative. A 10-case series. J Periodontol 71(5):752-759.

Zafiropoulos GG, Hoffmann O, Kasaj A, Willershausen B, Weiss O, Van Dyke TE (2007). Treatment of intrabony defects using guided tissue regeneration and autogenous spongiosa

138

alone or combined with hydroxyapatite/beta-tricalcium phosphate bone substitute or bovine- derived xenograft. J Periodontol 78(11):2216-2225.

Zhang F, Zhang W, Shi H (2013). [Research progress of decellularized matrix in tissue engineered trachea transplantation]. Zhongguo Xiu Fu Chong Jian Wai Ke Za Zhi 27(2):223- 226.

Zhao YH, Zhang M, Liu NX, Lv X, Zhang J, Chen FM et al. (2013). The combined use of cell sheet fragments of periodontal ligament stem cells and platelet-rich fibrin granules for avulsed tooth reimplantation. Biomaterials 34(22):5506-5520.

Zhou Y, Chen F, Ho ST, Woodruff MA, Lim TM, Hutmacher DW (2007). Combined marrow stromal cell-sheet techniques and high-strength biodegradable composite scaffolds for engineered functional bone grafts. Biomaterials 28(5):814-824.

Zhu W, Liang M (2015). Periodontal ligament stem cells: current status, concerns, and future prospects. Stem Cells Int 2015(972313.

Zuk PA, Zhu M, Mizuno H, Huang J, Futrell JW, Katz AJ et al. (2001). Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 7(2):211-228.

139

Appendix

Instruction manual for decellularized constructs fabrication and validation

1. Materials

All fluids used in the cell culture and decellularization processes have to be sterile and warmed to 37 oC. Aseptic techniques should be used at times.

o All cell culture incubation must be performed in a 37 C, 10% CO2 humidified incubator.

1.1. Primary Human periodontal ligament cell (hPDLC) harvesting and expansion

1. Human periodontal ligament tissue harvested from freshly extracted teeth (see

details in 3.1).

2. 25 and 175 cm2 culture flasks.

3. Dulbecco’s Modification of Eagle’s medium (DMEM).

4. 10% foetal calf serum (FCS). 5. Penicillin (50units/ml, Invitrogen). 6. Streptomycin (50μg/ml, Invitrogen). 7. 24 well culture plates. 8. Ascorbic acid (100 & 1000µg/mL). 9. Cell culture incubator. 10. Centrifuge machine. 11. 50 mL falcon tubes (sterile).

140

1.2. Melt electrospun PCL carrier membrane fabrication

12. Direct writing melt electrospinner with suitable G-coding software (custom made

machine was fabricated in-house at the Institute of Health and Biomedical

Innovation, Queensland University of Technology, (Brown et al., 2011)). 13. Medical grade Polycaprolactone (PCL).

1.3. Cell sheet fabrication and harvesting

14. Fine pointed curved tweezers (sterile). 15. 5 mm sterile dermal Biopsy punch (Kai Medical®). 16. PCL scaffolds with 0-90o fiber orientations.

1.4. Perfusion decellularization components

17. Phosphate buffered saline (PBS). 18. 0.05% Trypsin.

19. 20 mM NH4OH solution.

20. 0.5% v/v Triton X-100.

21. Bi-directional perfusion system (explained in section 2.4.).

22. DNase I solution (100 U/mL, Invitrogen, Cat. No. 18047-019).

23. CaCl2 (0.9 mM) and MgCl2 (0.5 mM) in sterile PBS.

24. Sterile double distilled water.

25. Petri dishes (sterile).

1.5. Cell sheet fixation and preparation for immunostaining for confocal imaging.

26. 4% paraformaldehyde solution in PBS at pH 7.4 (Sigma-Aldrich).

141

27. Triton X-100 (0.2%) in PBS.

28. Bovine serum albumin 1% (Sigma-Aldrich) in PBS.

29. Antibodies against human Collagen I and Fibronectin (Life Technologies,

Invitrogen).

30. 4',6-diamidino-2-phenylindole (DAPI, 5 µg/ml).

31. Phalloidin – tetramethylrhodamine B isothiocyanate conjugate (Phalloidin-

TRITC, 0.8 U/ml, Life Technologies, Invitrogen).

32. Fluorescently labelled secondary antibody Alexa 633 goat anti-mouse antibody (5

μg/mL, Alexa Fluor, Catalog #A-21126, Invitrogen).

1.6. Growth factor extraction

33. NaCl (2M).

34. 20 mM HEPES.

35. EDTA protease inhibitor cocktail (Roche complete mini, Roche Applied Science,

Indianapolis, IN).

2. Methods

All procedures should take place within a cell culture room in a biosafety cabinet and under aseptic conditions.

2.1. Primary Human periodontal ligament cell (hPDLC) harvesting and expansion

1. Place redundant freshly extracted teeth immediately in warm DMEM.

2. Hold the extracted tooth by the crown in order to avoid any damage to the

periodontal tissues.

142

3. Obtain periodontal ligament (PDL) tissues by gently scraping the tissues attached

to the middle third of the roots.

4. Dice the tissues into smaller portions (approximately 1 x 1 mm).

5. Using a plastic pipette, transport the diced PDL tissues to a 25 cm2 flask and wet

them with one or two drops of DMEM. Place the flask in the incubator upright for

5 minutes, then re-wet the diced tissues again and incubate for further 5 minutes

until the PDL tissues are firmly attached to the flask inner surface. Add 5 mL of

DMEM supplemented with 10% foetal calf serum (FCS), Penicillin (50units/ml)

and Streptomycin (50µg/ml), with culture medium changes every 48 hours until

the cells reach 80% confluency (See note 1).

6. Passage the cells by discarding the culture medium, rinsing with warm PBS twice,

and then adding sufficient 0.25% Trypsin to just cover the cell layer. Place the

flask in the incubator for 1 minute, and then check under an inverted microscope

for cell detachment. Collect the cells in 10 mL of DMEM and then split the cells

in a 1:3 ratio.

7. Use cells between the 3rd and 4th passages for optimal cell growth and expansion

(See note 2).

2.2.Melt electrospun PCL carrier membrane fabrication

1. Load medical grade polycaprolactone granules into a syringe, set the temperature

of the water tank to 100oC until the PCL melts completely (See note 3).

2. Adjust the spinneret collector distance to 2 cm.

3. Set the feed rate to 20 µL/hr and the voltage to 10kV.

4. Set the transitional speed of the collector at 250 mm/min in order to obtain straight

fibers.

143

5. Set the spinner to deposit alternating series of layers with 90 degree orientation.

6. Collect the deposited scaffolds using clean fine tipped tweezers into a clean Petri

dish and seal the dish with parafilm.

7. PCL scaffolds must be cut into 5 mm diameter using a biopsy punch and sterilized

before cell sheet harvesting (See note 4). Use UV in a biosafety cabinet to perform

scaffold sterilization overnight where scaffolds should be embedded in ethanol

100%.

2.3. Cell sheet fabrication and harvesting

1. Discard culture medium from the cell culture flask/s, rinse gently twice with warm

PBS then detach the cells by adding Trypsin as described 3.1.6.

2. Re-suspend the cells using a master mixture of 50 mL into a 24 well cell culture

plate, with seeding density of 50,000 cells/well.

3. Leave the cells in high glucose DMEM for the first 24 hrs.

4. After 1 day post seeding, discard the old medium and add 200 µL of DMEM

supplemented with 1000 mg/mL ascorbic acid for a further 72 hrs.

5. Discard the medium and add the same volume of DMEM but only with 100

mg/mL ascorbic acid, and change this medium every 48 hrs for the following 15

days until the cell sheets can be visibly detected .

6. Wet the 5 mm diameter PCL scaffolds with DMEM.

7. Place the PCL scaffold exactly in the center of each well after discarding most of

the medium from each well (See note 5).

8. Make sure that the scaffold is in contact with the cell sheet and avoid excessive

forces that may damage the cell sheet.

144

9. Using fine curved tweezers, start detaching the cell sheets from the boundaries of

the wells in a circumferential manner (see note 6).

10. Pull the cell sheet edges upwards and towards the center of each well, fold it over

the scaffold.

11. Gently use the tweezers to flip the scaffold with the cell sheet facing upwards.

12. Wet the detached cell sheet with 10-20 µL of medium every 5 minutes for 20

minutes total, with the culture plate placed back into the incubator after each

wetting cycle.

13. Transfer the cell sheet constructs (scaffold + cell sheet/ CSC) to a new 6 well cell

culture plate, then add 300 µL of DMEM/well (See note 7).

14. Leave the constructs overnight so the cell sheets adhere to the PCL scaffolds.

2.4. Cell sheet construct decellularization

1. Use a perfusion system for the cell sheet construct (CSCs) decellularization.

2. We designed a perfusion bioreactor system composed of an infusion/withdrawal

syringe pump, 30 mL plastic syringes, silicone tubes and decellularization

chambers fabricated from photo-curable material. The chambers and its

components were designed with CAD software and additive manufactured using

an inkjet 3D printer (Objet30 Pro Desktop, Stratasys) using an acrylic resin

(Verowhite Plus 835, Stratasys).

3. All components must be placed in a biosafety cabinet and exposed to UV

radiation overnight for sterilization.

4. Rinse the CSCs once with warm PBS at 37oC, and place them in the

decellularization chambers with a maximum of 11 constructs per chamber (See

note 8&9).

145

5. Perfuse the CSCs in 30 mL of 20 mM NH4OH solution with 0.5% v/v Triton X-

100 at room temperature (See note 10).

6. Bi-directional perfusion of the constructs needs to be performed for 60 min at a

rate of 1,000 mL/hr, with a flow inversion every 50s (See note 11).

7. Discard the liquids from the decellularization chambers by simply detaching the

silicone tubes from the perfusion chambers, then immediately connect another

loaded syringe for the DNase perfusion step to the perfusion chambers.

8. Perfuse with 30mL of DNase I solution (100 U/mL, Invitrogen) in CaCl2 (0.9

o mM) and MgCl2 (0.5 mM) in sterile PBS at 37 C for 60 min (See note 12&13).

9. Finally, perfuse the constructs with sterile water at 37oC for another 60 min.

10. Carefully disconnect the perfusion chambers from the syringe pump after

discarding all fluids.

11. Collect the CSCs by opening the chambers inside the cabinet using sterile

tweezers. Then place them in a culture Petri dish, add sterile PBS to completely

cover the scaffolds, seal the dishes completely with parafilm then place them in a

refrigerator overnight at 4oC.

2.5. Immunostaining of cell sheet constructs

1. Place constructs in a clean multi-well culture plate and rinse samples carefully

twice with PBS at room temperature.

2. Fix constructs in 4% paraformaldehyde (PFA) for 20-30 minutes.

3. Discard PFA then wash once with PBS.

4. If necessary, Permeabilize cells for 5 minutes in 0.2% Triton X-100 in PBS.

5. Discard solutions then wash samples twice gently with PBS.

6. Transfer samples into 1% BSA in PBS and incubate for 10 minutes.

146

7. Prepare solutions of antibodies diluted in 1% BSA in PBS, 1:200 for collagen type

I and 1:300 for fibronectin.

8. Add 100-200 µL of antibody solution for each well then incubate for 45 minutes.

9. Rinse constructs three times with PBS.

10. Incubate samples in 1% BSA in PBS containing fluorescently labeled secondary

antibody (Alexa 633 goat anti-mouse, 5 µg/mL), 0.8 U/mL TRITC-conjugated

phalloidin and 5 µg/mL DAPI for 45 minutes protected from light.

11. Rinse twice with PBS at room temperature then proceed with confocal imaging.

2.6. Growth factor extraction

1. Dissolve 1 tablet of the protease inhibitor in 10 ml of the extraction buffer solution

(2M NaCl in 20 mM HEPES).

2. Rinse constructs twice with PBS at room temperature.

3. Add 300 µL of extraction buffer solution with protease inhibitor for each

well/construct.

4. Seal the plate with parafilm and incubate for 60 minutes on an orbital shaker at

room temperature.

5. Collect the supernatant for each construct into a sterile eppendorf tube and store at

-80 oC.

3. Notes

1. Avoid cell over-confluency during the cell propagation (expansion) phase as it

affects their future survival, growth and ability to form extracellular matrix.

147

2. When seeding in large 175 cm2, start with at least 500,000 cells as lower seeding

densities are much slower to reach semi-confluency.

3. Multiple biomaterial options exist for the carrier component of the construct. We

utilized electro-spun polycaprolactone (PCL) because it is a medical grade material

with excellent biocompatibility. It has the key properties of being able to be

fabricated in a highly porous structure that allows perfusion of the decellularization

reagents, while still retaining appropriate mechanical properties to provide

sufficient support during the perfusion process. It is also a chemically stable

polymer that retains its structural integrity during the decellularization process.

Other biomaterials with similar properties could also be utilized as the cell sheet

carrier component of the construct.

4. If using PCL scaffolds as a carrier for the cell sheets, it is advisable to enhance the

scaffold hydrophilicity by immersion into 2M NaOH for 30 minutes followed by 5

rinses of ultrapure water.

5. Avoid excessive pressure when placing a scaffold on top of a cell sheet during the

cell sheet harvesting step.

6. When using fine tweezers to detach a cell sheet, always aim the tip to the

boundaries of the well, to be able to harvest an intact cell sheet.

7. When relocating or transferring the constructs from one plate to another or to

perfusion chambers, always handle the constructs with extra care holding each one

separately at a time. Use broad tipped tweezers for this procedure while holding

the constructs from the sides, also avoid squeezing the construct in order not to

damage the cell sheet and/or the scaffold.

8. Use spacers between the scaffolds inside the perfusion chamber.

148

9. Arrange the scaffolds with enough distance between each, so they do not adhere to

each other. Avoid overfilling of the decellularization chambers with fluids.

However, please note that scaffolds must be immersed completely in fluids all the

time.

10. Avoid rapid fluid perfusion as this will disrupt the constructs.

11. CaCl2 (0.9 mM) and MgCl2 (0.5 mM) concentrations in PBS are essential to

activate the DNase enzyme. However, if these concentrations are exceeded,

calcium ions will chelate with phosphate forming a turbid white solution. If this

occurs during the DNase perfusion step, the solution should be discarded and a

fresh one prepared.

12. DNase enzyme concentration and the period of perfusion should not exceed what

is outlined in point 3.4.8, as this will damage the extracellular matrix components.

149