Gene-Augmented Mesenchymal Stem Cells in Bone Repair

Home , Epicondyle

GENE-AUGMENTED MESENCHYMAL STEM CELLS IN BONE REPAIR

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Terri A. Zachos, DVM

* * * * *

The Ohio State University 2006

Dissertation Committee: Approved by Professor Alicia Bertone, Adviser

Professor Jeffrey Bartlett ______Adviser Professor Clifford Les Graduate Program in Veterinary Clinical Sciences

Professor Thomas Rosol

ABSTRACT

Complicated healing of articular fractures represents a clinical challenge and a financial burden on the health care system. Biologic repair systems are being evaluated, in both in vitro and in vivo experimental models, to augment healing of musculoskeletal connective tissues to address this problem.

Bone marrow-derived mesenchymal stem cells (BMDMSC) hold promise for targeted osteogenic differentiation and can be augmented by delivery of genes encoding bone morphogenetic proteins (BMP).

The feasibility of promoting osteogenic differentiation of BMDMSC was investigated using two BMP genes in monolayer and three-dimensional alginate culture systems. Cultured BMDMSC were transduced with E1-deleted adenoviral vectors containing either human BMP2 or BMP6 coding sequence under cytomegalovirus (CMV) promotor control and either sustained in monolayer or suspended in 1 ml 1.2% alginate beads for 22 days. Adenovirus (Ad)-BMP-2 and Ad-BMP-6 transduction resulted in abundant

BMP-2 and BMP-6 mRNA and ligand expression in monolayer culture and BMP-2 ligand expression in alginate cultures. Ad-BMP-2 and Ad-BMP-6 transduced BMDMSC in monolayer had earlier and robust alkaline phosphatase-positive staining and mineralization and were sustained for a longer duration with morphology scores more consistent with viable cells than untransduced or Ad-ß-galactosidase-transduced cells. Ad-BMP-2- and to a lesser degree Ad-BMP-6-transduced BMDMSC suspended in alginate demonstrated greater mineralization than untransduced cells. Gene expression studies at day 2 confirmed an inflammatory response to the gene delivery process with up-regulation of interleukin 8 and CXCL2. Up- regulation of genes consistent with response to BMP exposure and osteogenic differentiation, specifically endochondral ossification and extracellular matrix proteins, occurred in BMP-transduced cells. These data support that transduction of BMDMSC with Ad-BMP-2 or Ad-BMP-6 can accelerate osteogenic differentiation and mineralization of stem cells in culture, including in three-dimensional culture. BMP-2- transduced stem cells suspended in alginate culture may be a practical carrier system to support bone formation in vivo. BMP-6 induced a less robust cellular response than BMP-2, particularly in alginate.

In vivo models are valuable to evaluate fracture healing methods. A weight-bearing, distal femoral

intercondylar articular osteotomy model was created in the nude rat. Osteotomies were treated with

BMDMSC, either wild-type (NoAd) or transduced with an adenoviral-bone morphogenetic protein 2

transgene construct (Ad-BMP-2). Cells were delivered in alginate (ALG) or injected in saline. Controls

were empty ALG, saline injections, direct Ad-BMP-2 injection, and untreated osteotomies. Healing was

compared using quantitative micro-computed tomography, fluorescent labeling, and histology. At day 14,

osteotomy gap area in the Ad-BMP-2 ALG group was significantly greater than any other group (P <

0.0003). The group treated with Ad-BMP-2-transduced cells injected in saline (Ad-BMP-2 cells) had healed with less osteotomy gap area (P < 0.0001) and volume (P < 0.02) than untreated controls. In ALG groups, bone healing was impeded by the development of a chondroid mass most pronounced in the Ad-

BMP-2 ALG group. Injection of Ad-BMP-2-transduced BMDMSC in saline accelerated bone healing and reconstituted the articular surface in this distal femoral osteotomy model of articular fracture healing.

iii

Dedicated to my parents, Evelyn Stasinopoulos Zachos and George H. Zachos, who

taught me that “with God all things are possible” (Matthew 19:26).

ACKNOWLEDGMENTS

I sincerely thank my adviser, Dr. Alicia Bertone, for providing exceptional mentorship, advice, and support, and for setting an example that has challenged me to strive for excellence in every endeavor. I thank Dr. Clifford Les for extensive discussions and consultation in the areas of imaging, statistical analysis, and biomechanics, and for his valuable advice on grantsmanship and collaboration. I thank Dr.

Les and Drs. Jeffrey Bartlett and Thomas Rosol for their tireless efforts as members of my Doctoral

Advisory, Candidacy Examination, and Final Oral Examination, and Dissertation Committees. I gratefully acknowledge Tim Vojt for electronic artwork and animations. I thank Alan Bakaletz, Ruth Berger, Joseph

Ielapi, Amanda Johnson, Susie Jones, Dr. Yi-wen Liu-Stratton, and Anne Saulsbery for technical support, and Dr. Kelly Santangelo for managing local anesthesia in in vivo studies. I am grateful for the assistance of Dr. Valerie Samii and Linnea Baumwart with computed tomography in the Veterinary Teaching

Hospital. I thank Dr. Steven Weisbrode for histopathologic review and consultation. I am indebted to

Alisha Diggs, without whom micro-computed tomography studies would not have been possible, and to

Jessica Williams, Colleen Flanagan, and Dr. Scott Hollister for assistance with micro-computed tomography data analysis. Some of the materials employed in this work were provided by the Tulane

Center for Gene Therapy through a grant from NCRR of the NIH, Grant # P40RR017447. I thank Roxanne

Reger, Margaret Wolfe, and Dr. Darwin Prockop of the Tulane Center for Gene Therapy for proving mesenchymal stem cells and their experience on their biological behavior in vitro.

My graduate program and this research were supported by a Ruth L. Kirschstein Individual

National Research Service Award from the National Institutes of Health/National Institute of Arthritis and

Musculoskeletal and Skin Diseases (Grant # F32AR050916) and by the Trueman Endowment.

VITA

November 4, 1970 ...... Born in Flushing, NY

1988 ...... Diploma with Honors, The Peddie School, Hightstown, NJ

1992 ...... BS with Honors, Cornell University

1996 ...... DVM, Cornell University

1996-1997 ...... Intern, Small animal medicine and surgery The Animal Medical Center New York, NY

1997-2000 ...... Resident, Small animal surgery The Animal Medical Center New York, NY

2000-2002 ...... Sten-Erik Olsson Fellow in Comparative Orthopaedic Research Michigan State University

2002-2003 ...... Graduate Research Associate Department of Veterinary Clinical Sciences The Ohio State University

2003-2006 ...... Graduate Fellow Department of Veterinary Clinical Sciences The Ohio State University

PUBLICATIONS

1. Munsterman AS, Bertone AL, Zachos TA, Weisbrode SE. Effects of the omega-3 fatty acid, alpha-linolenic acid, on lipopolysaccharide-challenged and -unchallenged equine synovial explants, American Journal of Veterinary Research, 66: 1503-1508, 2005.

2. Zachos TA, Bertone AL. Growth factors and their potential therapeutic applications for healing of musculoskeletal and other connective tissues, Am J Vet Res, 66: 727-738, 2005.

FIELDS OF STUDY

Major Field: Veterinary Clinical Sciences

TABLE OF CONTENTS

Page

Abstract ...... ii Dedication ...... iv Acknowledgments ...... v Vita ...... vi List of Tables ...... x List of Figures ...... xi

Chapters:

1. Growth factors: interactions and therapeutic implications for musculoskeletal connective tissues ...... 1 1.1 Summary ...... 1

1.2 Musculoskeletal and connective tissues: clinical concerns...... 1

1.3 Review of terms...... 3

1.4 The biology of growth factors ...... 6

1.5 Biologic functions and signaling interactions among growth factors in musculoskeletal and connective tissues...... 8

1.6 Molecular interactions relevant to musculoskeletal and connective tissue development and healing ...... 17

1.7 Perspectives on the clinical applications of growth factors, disease-modifying agents and biologics ...... 19

1.8 Clinical implications for new therapies and devices...... 20

vii

1.9 Clinically available therapies and devices...... 23

1.10 The clinical future for therapies and devices ...... 26

1.11 Conclusions ...... 30

2. Gene-mediated osteogenic differentiation of stem cells by BMP2 or BMP6 . . . .32

2.1 Summary ...... 32

2.2 Introduction ...... 33

2.3 Methods ...... 36

2.4 Results ...... 41

2.5 Discussion ...... 50

3. Rodent models for the study of articular fracture healing ...... 58

3.1 Summary ...... 58

3.2 Introduction ...... 59

3.3 Methods ...... 60

3.4 Results ...... 68

3.5 Discussion ...... 71

4. Chondro-osseous differentiation of bone marrow-derived mesenchymal stem cells in alginate cultures induced by BMP-2 and BMP-6 gene delivery ...... 81

4.1 Summary ...... 81

4.2 Introduction ...... 82

4.3 Results ...... 84

4.4 Discussion ...... 86

4.5 Methods ...... 90

viii

5. Mesenchymal stem cell-mediated gene delivery of BMP-2 in an articular fracture model ...... 102

5.1 Summary ...... 102

5.2 Introduction ...... 103

5.3 Results...... 106

5.4 Discussion...... 110

5.5 Methods...... 115

Appendix A: Microarray Data From In Vitro Studies ...... 139

Appendix B: Supplementary Micro-Computed Tomography Data ...... 146

Bibliography ...... 153

LIST OF TABLES

Table Page

1.1 Classification, origin, and major functions of growth factors relevant to musculoskeletal and connective tissue healing ...... 7

1.2 Dose of rhBMP and rate of bone healing in various species in the presence of recombinant bone morphogenetic proteins (BMPs) ...... 22

1.3 Commercially available growth factors and their respective costs (as quoted for purchase on a unit basis) ...... 24

2.1 Transgene protein expression (mean ± SEM) of bone marrow-derived mesenchymal stem cells ...... 44

3.1 Evaluation time points for mice and rats ...... 80

 4.1 Custom-designed primers and probes for Taqman real-time RT-PCR ...... 97

4.2 Quantitative parameters used to evaluate cell-alginate and empty alginate constructs ...... 98

5.1 Values for parameters measured using micro-computed tomography (micro-CT) images ...... 126

A1 Genes differentially regulated (>3-fold change) in response to transduction with Ad-BMP-2, Ad-BMP-6, or Ad-Luc ...... 140-144 A2 Genes relevant to osteogenic and/or chondrogenic differentiation whose expression was not altered by adenoviral transduction of hbmp2 or hbmp6 at day 2 or day 12 ...... 145

LIST OF FIGURES

Figure Page

1.1 Generalized mechanism of regulation of gene expression and translation into proteins as mediated by growth factors...... 5

1.2 Subfamily classifications within the bone morphogenetic protein (BMP) family of proteins...... 10

1.3 Interactions between bone morphogenetic proteins (BMPs), Indian hedgehog protein (Ihh), parathyroid hormone-related peptide (PTHrP), parathyroid hormone-related peptide receptor, Smad intracellular transcription factors, and the transcription factor Runx2 in regulation of chondrocyte maturation in the growth plate ...... 20

2.1 Appearance of equine BMDMSC in monolayer (upper two rows) and alginate (lower two rows) culture ...... 43

2.2 Detection of luciferase in equine BMDMSC in monolayer cultures and alginate constructs ...... 45

2.3 Representative agarose gel (1.2%) electrophoresis of real-time RT-PCR products ...... 47 2.4 Comparisons of von Kossa and X-gal staining of alginate and monolayer cultures ...... 48

3.1 Illustration of the rat stifle joint depicting the lateral intercondylar osteotomy model...... 75

3.2 Luciferase expression in distal femora of mice ...... 76

3.3 Faxitron high-detail radiographs taken of mouse stifle joints at days 0 and 21. 77

3.4 Faxitron high-detail radiographs taken of two stifle joints from athymic nude rats (NIH rnu) at day 14...... 77

3.5 Axial computed tomographic (CT) images of healing intercondylar osteotomies in NIH rnu (athymic nude) rats at day 21...... 78

3.6 Proton density-weighted magnetic resonance images (MRI) of healing intercondylar femoral osteotomy at day 21 in a mouse...... 78

3.7 Proton density (PD)-weighted MRI of unoperated control NIH rnu rat stifle joint obtained with 4.7T magnet showing trabecular bone, physes, articular cartilage, and articular and periarticular soft tissue structures...... 79

3.8 Light microscopic appearances of medial intercondylar femoral osteotomies in SKH1 (immunocompetent) mice...... 79

3.9 Light microscopic appearances of lateral intercondylar femoral osteotomies in NIH rnu (athymic nude) rats on day 14...... 80

4.1 Median size of cell-alginate constructs from each treatment group. Bars represent median values; error bars demonstrate range...... 99

4.2 Cell-alginate constructs at day 22, with hematoxylin and eosin, toluidine blue, and von Kossa staining (X400)...... 100

4.3 Immunostaining of cell-alginate constructs for type II collagen at day 22 (X400) ...... 101

5.1 Expression of luciferase, quantified as counts (numerical photon data count shown with a pseudocolor display [http://www.xenogen.com/wt/page/software]), in femora of rats receiving alginate constructs with cells transduced with adenoviral luciferase gene (AdLuc) into an articular osteotomy ...... 127

5.2 Two-dimensional micro-computed tomographic (micro-CT) images (GEHC Microview) in the coronal, sagittal, and axial (transverse) imaging planes (from left to right), demonstrating differences in appearances of osteotomy gaps among treatment groups ...... 128-130

5.3 Selected isosurface renderings of micro-computed tomographic images ...... 131

5.4 Osteotomy gap area values calculated from micro-computed tomographic (micro- CT) images at day 14 ...... 132

5.5 Histology of decalcified sections, stained with safranin O/fast green (X400), representative of each treatment group 14 days after treatment ...... 133

5.6 Comparison of histologic appearance of articular surfaces ...... 134 xii

5.7 Calcified sections: light microscopic appearance of lateral femoral condyles of rats ...... 135

5.8 A schematic, drawn from radiographs of the rat stifle joint, depicting the lateral intercondylar osteotomy (dotted line) performed in athymic nude rats ...... 136

5.9 Lateral intercondylar osteotomy of the distal femur in a 180-gram nude rat . . .137

5.10 Use of the In Vivo Imaging System (IVIS®) on live bone marrow-derived mesenchymal stem cells (BMDMSC) in vitro genetically modified to express the reporter gene product luciferase...... 137

5.11 Use of the In Vivo Imaging System (IVIS®, Xenogen Corporation, Alameda, CA) to detect luciferase reporter gene product expression by genetically modified bone marrow-derived mesenchymal stem cells (BMDMSC) in vivo...... 138

B1 Representative two-dimensional micro-computed tomographic (micro-CT) images in the coronal, sagittal, and axial (transverse) imaging planes demonstrating differences in appearances of osteotomy gap among treatment groups ...... 147-149

B2 Representative isosurface renderings of micro-computed tomographic (micro-CT) images demonstrating cranial, caudal, and craniolateral unhealed bone in the Untreated (untreated osteotomy), EMPTY ALG (alginate without cells), NoAd ALG (untransduced bone marrow-derived mesenchymal stem cells [BMDMSC] suspended in alginate), AdBMP2 ALG (BMDMSC transduced with AdBMP2 and suspended in alginate), SALINE (saline injection without cells), and NoAd cells (unstransduced BMDMSC injected in saline suspension), groups ...... 150-152

xiii

CHAPTER 1

GROWTH FACTORS: INTERACTIONS AND THERAPEUTIC IMPLICATIONS FOR MUSCULOSKELETAL CONNECTIVE TISSUES

1.1 Summary

Growth factors are a diverse group of proteins that act as molecular signals for various cell types and promote generation, regeneration, and remodeling of various tissues and organs. Most research and development efforts have focused on the musculoskeletal, skin, and other connective tissues. The evolution of purification of recombinant proteins, and the sequencing of the human, mouse, rat, canine, and other genomes has expanded the potential for the use of growth factors as protein or molecular therapy for musculoskeletal, skin, and connective tissue healing.

1.2 Musculoskeletal and connective tissues: clinical concerns

The consideration of healing of connective tissues involves not only the connective tissues of the musculoskeletal system, but also of the skin, blood vessels, and eyes. What these diverse organ systems have in common are their collagenous extracellular matrices. The majority of investigations related to healing of

musculoskeletal and other connective tissues have focused on bone, diarthrodial joints, tendons and ligaments, and skin. While the potential indications for the use of growth factors in clinical settings are broad and diverse, this discussion will be centered on potential implications for healing of musculoskeletal connective tissues.

Clinical concerns pertinent to veterinary patients relating to bone healing are numerous. Among the most prominent of these is fracture healing. Of particular concern are cases of delayed and nonunion. Additional potential applications for augmentation of bone healing are conditions resulting in lysis of bone, which include fungal and bacterial osteomyelitis, metastatic disease and aseptic loosening of prostheses used in total joint replacement [1-4].

The fact that articular cartilage is avascular and aneural makes it among the most difficult tissues in the body to heal. Extensive research in the last five years has focused on this clinical concern in animals and humans, resulting in more than 750 citations in the scientific literature. Cartilage injury can occur as the result of acute trauma, chronic joint instability, or secondary to other chronic orthopedic disease. In veterinary patients, diseases of joints include osteochondral defects, (both those of traumatic origin as well as those secondary to developmental orthopedic disease) [5-9], osteoarthritis [10-17], and inflammatory arthritidies [18-20]. Osteoarthritis, both secondary to these osteochondral injuries, as well as as a primary disease entity, is estimated to affect 20% of the canine population over one year of age, and is one of the most common sources of chronic pain treated by veterinarians, according to a recent industry survey [21]. A recent study found radiographic evidence of osteoarthritis in 90% of cats greater than 12 years of age [22].

Injuries to the musculoskeletal soft tissues are also of great clinical significance to both veterinary and human patients and are the slowest to heal and regain biomechanical properties of any tissue. One of the most common ligament injuries in human and canine patients is rupture of the anterior cruciate ligament (ACL) [23].The incidence of this injury in people in the United States is estimated to be 1 in 3000, with approximately

95,000 new injuries occurring each year [24]. Acute and chronic injuries to tendons, particularly tendinosis of the Achilles, rotator cuff, elbow, hip, and quadriceps, are increasing in frequency [25]. The incidence of injury to the superficial digital flexor tendon of Thoroughbred racehorses is estimated to be between 8 and 35%, with only 20-

60% of these patients returning to athletic activity following treatment [26,27]. Active searches for growth factor therapies are in progress [28].

1.3 Review of terms

Potential therapeutic indications for growth factors will use recombinant proteins or molecular gene therapy. Recombinant proteins are proteins produced using recombinant deoxyribonuclease (DNA) technology to generate high production by cells, using a specific, identified gene sequence for the protein and subsequent purification of the proteins for use. Recombinant DNA technology permits the isolation of specific sequences of interest using restriction nucleases, DNA cloning to amplify copies of the same gene sequence, subcloning the sequences into high-production expression vectors and cells, and mass purification of a uniform protein for manufacturing and clinical use.

Alternatively, the gene sequence could be subcloned into an expression vector (modified

virus or plasmid) and this construct administered as a gene therapy to produce the protein by our own body’s cells [29]. Gene therapy has been defined as, “the science of transfer of genetic material into individuals for therapeutic purposes by altering cellular function or structure at the molecular level” [30]. The delivery of genetic material can be accomplished by one of two methods: direct (in vivo) gene delivery [31,32] and indirect

(ex vivo) gene delivery [30]. Direct gene delivery involves the delivery of the genetic material of interest into the target cells in vivo. Indirect gene delivery involves the genetic manipulation of cells under in vitro conditions and subsequent delivery to the desired anatomic site in the recipient. While direct gene delivery is less technically demanding and potentially less costly, it may result in lower transduction efficiency [33]. Indirect gene delivery has the advantage of allowing for a specific cell type to be targeted for genetic modification, and potentially expanded in vitro and implanted into the recipient at the time of maximum production of the desired protein product [34]. The main disadvantage associated with indirect gene delivery is the increased equipment, time, and cost, associated with the harvesting and processing of the target cells [33]. Other potential molecular therapies include “knock-down” of gene expression in our own body’s cells by administering segments of DNA that can regulate or block gene expression, termed antisense therapy or molecular therapy. While various growth factor proteins differ in their specific mechanisms of action based on their structure and target cells, a generalized mechanism by which these proteins modify gene expression and ultimately, translation into the proteins of interest, is given in Figure 1.1.

Figure 1.1: Generalized mechanism of regulation of gene expression and translation into proteins as mediated by growth factors. The growth factor is shown binding to a generalized transmembrane receptor. The generalized role of a transcription factor is shown.

This mechanism can be used in factories to generate commercial volumes of recombinant protein for manufacturing. A similar mechanism can be used to target a cell receptor for a vector containing a gene encoding a growth factor that can be transcribed and the protein produced in vivo.

1.4 The biology of growth factors

Growth factors with potential clinical applications in augmentation of healing of musculoskeletal and connective tissues will be categorized based on their biological activities, and include the transforming growth factor-β (TGF-β) superfamily. This group includes transforming growth factor-beta (TGF-β), the bone morphogenetic proteins

(BMPs), and the growth and differentiation factors (GDFs), activins, inhibins, and the

Müellerian substance [35,36]. Other growth factors with demonstrated potential for use in connective tissue healing include the insulin-like growth factors (IGFs), platelet-derived growth factor (PDGF), fibroblast growth factors (FGFs), and epidermal growth factor

(EGF). These proteins, their natural sources and primary functions are listed in Table 1.1.

The identification and localization of the role of growth factors in skeletal and tissue development has opened an area of medicine to develop these proteins for use as disease-modifying agents. For example, it has been proposed that fracture healing recapitulates the process of endochondral ossification that occurs in the development of the embryonic limb [37-41]. The experimental evidence confirming this hypothesis has triggered tremendous efforts to determine the potential use of the growth factors involved in this process in clinical settings to augment compromised healing of connective tissues.

Growth Factor Origin Major Function (s) Transforming Growth Extracellular matrix Initiates differentiation of pluripotential Factor-β (TGF-β) of cartilage and mesenchymal cells; triggers proliferation of bone; platelets resulting chondrocytes and osteoblasts; regulates synthesis of collagen and proteoglycans Bone morphogenetic Extracellular matrix Initiate bone formation by differentiation of proteins (BMPs) of bone; osteoblasts; pluripotential mesenchymal cells into osteoprogenitor cells chondrocytes and osteoblasts; initiates differentiation of osteoprogenitor cells into osteoblasts; influences multiple aspects of skeletal patterning and development Insulin-like growth Chondrocytes, Induce proliferation and differentiation of factors (IGFs) osteoblasts, osteoprogenitor cells extracellular matrix of bone Platelet-derived growth Platelets, endothelial Stimulates proliferation of osteoblasts and factor (PDGF) cells, monocytes chondrocytes; acts as a mitogen for osteoblasts and pluripotential mesenchymal cells; is chemotactic for macrophages Fibroblast growth Osteoblasts, Stimulates growth of osteoblasts, chondrocytes, factors (FGFs) chondrocytes, and mesenchymal cells; induces replication of macrophages and inflammatory cells; stimulates differentiation of other inflammatory pluripotential cells; plays a role in angiogenesis cells, mesenchymal cells Epidermal growth Epithelial cells and Induces development of endothelial cells; factor (EGF) fibroblasts stimulates angiogenesis

Table 1.1: Classification, origin, and major functions of growth factors relevant to musculoskeletal and connective tissue healing. (Modified from Lieberman et al .J Bone Joint Surg 2002; Truumees and Herkowitz U Penn Orth J 1999); and Lind. Acta Orthop Scand 67(4):407-417, 1996).

A large volume of data has recently been published in the human orthopedic literature detailing the results of in vitro studies, in vivo studies, and human clinical trials employing recombinant growth factors and tissue-engineered scaffolds to augment fracture healing and spinal fusion. While the number of reports of clinical use of commercially available growth factors is limited, all of the studies preceding clinical

trials on human patients required detailed descriptions of the biological behavior of these substances in animal models. In light of the success associated with their use in experimental settings, there is tremendous potential to expand the clinical use of commercially available growth factors in veterinary patients.

Although there is abundant evidence in the literature documenting the potential for the use of growth factors in the healing of musculoskeletal connective tissues, there is also evidence to justify concerns associated with the over expression of certain growth factors and their receptors. In particular, the expression of BMP receptors has been shown to correlate with incidence of metastasis in osteosarcoma, suggesting that the

BMP pathway may play a role in the evolution of an aggressive biological behavior in this tumor. In addition, BMP receptors have been identified in other sarcomas [42,43].

These findings emphasize the need for judicious use of these potential mechanisms in the attempted augmentation of healing.

1.5 Biologic functions and signaling interactions among growth factors in

musculoskeletal and connective tissues

Transforming growth factor-β (TGF-β) is a protein found in platelets, bone, and cartilage. While the presence of TGF-β receptors on osteoblasts [44] and chondrocytes

[45] suggests a role in fracture healing, its exact role is unclear and may be unpredictable, depending on dose, species, and/or biological environment. Transforming growth factor-

β induces BMP expression, a more specific downstream regulator of bone formation,

which has received extensive study and clinical application. Studies in rabbit and rat models have demonstrated fracture healing with TGF-β treatment; however very high doses were used [46,47]. Other data from similar studies [48] conflict with these findings.

As a result, there are no known clinically applicable formulations of TGF-β at this time.

Among the most widely publicized studies are those utilizing recombinant human bone morphogenetic proteins (rhBMPs). The BMPs are members of the transforming growth factor-β (TGF-β) superfamily of secreted proteins. The BMPs are dimeric molecules comprised of two polypeptide chains and a single disulfide bond [49-51]. As more of these proteins have been identified and characterized in recent years, the BMP family has been further divided into a number of BMP subfamilies, recently summarized by Reddi [51] (Figure 1.2).

A dramatic change in approaches to musculoskeletal and connective tissue healing evolved in the late 1980’s and early 1990’s with the publication of several landmark studies initiated in the 1970’s. Triggered by the discovery of the bone morphogenetic proteins in Urist’s classic publication, “Bone: Formation by

Autoinduction,” in Science in 1965 [52], the concept that agents produced by the body itself could be used to treat injury and disease emerged. The investigations which ensued in the next twenty years documenting that proteins derived from demineralized bone matrix induced ectopic and orthotopic bone formation coincided with the cloning and amino acid sequencing of these proteins and subsequently, widespread application of recombinant DNA technology [53-55]. Urist’s discovery of BMPs in 196552 documented

Figure 1.2: Subfamily classifications within the bone morphogenetic protein (BMP) family of proteins.. The names of BMP subfamilies are shown in ovals. Alternative names for each protein are given in parentheses. (Modified from Reddi AH. Bone morphogenetic proteins: from basic science to clinical applications. J Bone Joint Surg Am 83 Suppl 1:S1-S6, 2001)

that the implantation of demineralized bone matrix induced bone formation in muscle

pouches in rats. Urist called this production of ectopic bone in vivo the bone induction principle. Determination of the sequence of the gene encoding human bone morphogenetic protein [54] has enabled the development of the recombinant protein, allowing tremendous expansion of its use in the enhancement of bone healing. Preclinical studies, performed both in vitro and on experimental animals, have yielded a large volume of data indicating the safety and efficacy of the use of recombinant BMPs in clinically relevant canine models, as well as in other species. Canine studies have focused primarily on three areas: healing of defects in long bones, spinal fusion, and dental and

maxillofacial defects. Heckman et al [56] demonstrated healing in a canine radial nonunion model with the implantation of canine recombinant BMP2 in a polylactic acid

(PLA) carrier. In similar critical-sized defects in canine radii, Sciadini et al [57] demonstrated increased new bone formation and superior biomechanical strength in radii implanted with a bovine-origin bone protein in a natural coralline calcium carbonate carrier reconstituted with bovine type I collagen. Sciadini and Johnson [58] established a dose-response relationship and a relative optimal dose for the use of rhBMP2 in a canine radial defect model. The results of this study suggested that rhBMP2 delivered in a collagen sponge is a safe alternative to autogenous cancellous bone grafting in diaphyseal defects in the canine radius.

Pluhar et al [59] and Zabka et al [60] evaluated the use of rhBMP2 in an absorbable collagen sponge carrier in the augmentation of host bone-allograft junctions in a canine intercalary femoral defect model using frozen allograft bone stabilized with a statically locked intramedullary nail. Use of rhBMP2 at this dose yielded results equal or superior to those obtained using autogenous cancellous bone graft, as evaluated by serial radiography, dual energy X-ray absorptiometry scans, gait analysis, biomechanical testing, and histopathologic evaluation.

Studies demonstrating the efficacy of rhBMP2 in achieving safe, rapid, consistent interspinous process [61] or intertransverse process [62,63] fusions in canine models have also been demonstrated. Recombinant human BMP2 has also been used in an absorbable collagen sponge (ACS) carrier to augment peri-implant osteogenesis in the integration of dental implants [64,65]. While these models are less likely to have clinical applications in

small animal surgery, they provide valuable data confirming the safe, rapid production of bone in vivo in response to treatment with surgically implantable formulations of rhBMP2.

In the first study evaluating the healing of tendon to bone using BMPs, Rodeo et al [66] examined the use of rhBMP2 in an absorbable type I collagen sponge carrier using two different doses to heal the long digital extensor tendon transplanted into a bone tunnel in adult dogs. The above treatments were compared to treatment with the collagen sponge carrier alone. Superior pullout strength and evidence of healing based on histopathologic analysis were found in the group treated with the lower concentration of rhBMP2. These results demonstrate that rhBMP2 can be used to augment tendon to bone healing in dogs. These findings have numerous potential clinical applications in small animal surgery, including but not limited to cranial cruciate ligament reconstruction and reconstruction of periarticular soft tissue structures in traumatized joints. In addition, previous findings documented that the beneficial effects of rhBMP2 treatment are decreased with doses exceeding an apparent threshold. In summary, the literature is replete with in vivo canine studies defining the dose, carrier systems, biocompatibility, and efficacy of rhBMP2 in canine bone in numerous applications involving long bones defects, spinal fusion, and reconstructive oral and maxillofacial surgery.

Cook et al first documented the efficacy of recombinant human osteogenic protein-1 (rhOP-1, also known as rhBMP7) in the healing of segmental osteoperiosteal defects in canine ulnae [67]. Other studies have documented augmented healing of bone in canine models using rhOP-1. These include the demonstration of improved and

accelerated healing of strut allografts to the femur [68]. In addition, Salkeld et al showed that rhOP-1 in a type I bovine collagen carrier can be used with allograft bone, with or without autograft bone, to reconstruct critical-sized defects in canine ulnae [69]. This study demonstrated the potential flexibility afforded with the use of rhOP-1 for the reconstruction of large bone defects, which has considerable potential clinical applications, in both small animal and large animal orthopedics. Use of autograft, in addition to rhOP-1 and allograft bone, was not significantly different, in terms of the radiographic, biomechanical and histologic properties of the new bone formed at 12 weeks postoperatively. Recombinant human OP-1 has also been used in an induced model of osteonecrosis of the femoral head in dogs [70]. The use of rh-OP1, in addition to autogenous iliac crest strut grafts and cancellous bone chips, resulted in an increase in rate of healing of femoral head defects, based on radiographic evaluation, when compared with defects filled with autograft bone alone. In a femoral intercalary defect model similar to that used by Pluhar et al, described above, Cullinane et al [71] reported results similar to those of Pluhar et al [59], but with the use of rhOP-1 at the host- allograft interface. While greater callus area was noted radiographically in rh-OP1-treated grafts when compared with untreated grafts, differences were noted between the two groups in neither weight bearing nor biomechanical properties of postmortem specimens.

Platelet-derived growth factor (PDGF) is released by platelets during the initial stages of fracture healing and has been identified in healing fractures in both mice [72] and humans. [35,73] In addition, PDGF has demonstrated a mitogenic effect on osteoblasts in vitro. [76] The use of PDGF has been pursued commercially in the form of

platelet concentrate and will be discussed further in the Clinical Future for Therapies and Devices section.

Growth hormone and insulin-like growth factor (IGF) are two growth factors with related functions. Growth hormone stimulates bone growth. It is secreted by the anterior lobe of the pituitary gland and exerts its endocrine effect on target cells in the growth plates and liver. These target cells then release IGF [35,72]. Two IGFs have been identified: IGF-1 and IGF-2. While IGF-2 is the growth factor present in the greatest proportion in bone, IGF-1 is the more potent of the two, and has been identified in healing fractures [35,73,74]. Investigations into the mechanisms of action of IGF-1 in articular cartilage have demonstrated an anabolic effect on chondrocyte metabolism [75-

77]. These effects have been well demonstrated in equine chondrocytes. In equine cartilage explants conditioned with interleukin-1 (IL-1), a known mediator of inflammation, treatment with corticosteroids and IGF-1 in combination had a protective effect which prevented proteoglycan degradation seen in explants treated with corticosteroids without IGF-1, as well as an anabolic effect, resulting in an increase in collagen content [78,79]. In equine chondrocytes grown in fibrin discs, supplementation with IGF-1 resulted in increased matrix synthesis with maintenance of chondrocyte phenotype. In addition, a dose-response effect of IGF-1 on matrix synthesis was found

[80]. Treatment of equine bone marrow-derived mesenchymal stem cells (BMDMSC) in monolayer culture pretreated with TGF-β prior to treatment with IGF-1 resulted in a two- fold increase in proteoglycan content of the culture medium. When these cells were evaluated in three-dimensional culture, pretreatment with TGF-β, followed by treatment

with IGF-1, resulted in a significant increase in proteoglycan production and type II

procollagen mRNA levels [81]. Finally, the anabolic effect of IGF-1 on equine

chondrocytes in three-dimensional culture has been maintained and augmented in the

presence of supraphysiologic concentrations of IGF-1 [82].

In an equine in vivo model of full-thickness cartilage loss, autogenous fibrin clots enhanced with recombinant human (rh) IGF-1 resulted in significant increases in the proportion of type II collagen at six months postoperatively, when compared with that present in control defects filled with fibrin clots alone [83]. It was later shown in a similar in vivo equine model that IGF-1 gene expression did not correlate with increases in levels of type II collagen and aggrecan mRNA which accompanied the healing of full-thickness articular cartilage defects [84]. However, the repair of full-thickness articular cartilage defects in horses with chondrocyte-fibrin scaffolds augmented with IGF-1 resulted in improved filling of defects at eight months postoperatively, with greater type II collagen production in defects receiving IGF-1-treated scaffolds when compared with defects filled with chondrocyte-fibrin scaffolds without IGF-1 treatment [85]. These in vitro and

in vivo equine studies confirm the anabolic effect of IGF-1 on articular cartilage, as well

as demonstrating the protective effect of this growth factor, allowing for the maintenance

of cartilage matrix integrity in the presence of inflammatory mediators. However, the

exact molecular mechanisms underlying these relationships have yet to be determined.

Data on the effects of IGF-1 in canine osteoarthritis are limited. Recently, in a

study evaluating levels of IGF-1 and IGF-2 following rupture of the cranial cruciate

ligament and subsequent surgical stabilization of affected joints in canine patients, levels

of both IGF-1 and IGF-2 were elevated when compared with contralateral control

(uninjured) joints. In this study, while the levels of both IGFs decreased over time following surgical stabilization, levels remained higher than those of control joints for the duration of the study (9 months postoperatively) [86]. This study also evaluated levels of insulin-like growth factor binding proteins (IGFBP), demonstrating increased levels of two IGFBPs in injured joints, proposing that these proteins may play a role in limiting the availability of IGF-1 and IGF-2 in injured joints [87]. These studies clearly show the numerous complexities involved in the IGF-1 axis, suggesting that the treatment of articular cartilage defects with scaffolds containing a single growth factor (IGF-1) may not maximize the benefits potentially achievable by addressing cartilage healing with multiple components of the IGF axis, in addition to pretreatment with TGF-β. Despite this documented activity of IGF-1, there is insufficient data in the literature at this time to support its use in augmenting bone and cartilage healing.

One of the subfamilies of growth factors in the BMP family, the cartilage-derived morphogenetic proteins (CDMPs) is a distinct group of proteins also referred to as growth/differentiation factors (GDFs). This nomenclature has led to some confusion, as according to the molecular events occurring during bone formation, all BMPs are, strictly speaking, cartilage morphogenetic proteins, as cartilage differentiation, chondrocyte hypertrophy, and apoptosis precede bone formation [51]. The names of the members of the CDMP subfamily, as well as their BMP-specific names, are given in Figure 1.2.

In addition to the above studies utilizing BMPs, Millis et al [88] evaluated the effects of recombinant canine somatotropin (growth hormone) in critical-sized defects in

canine radii. This study documented superior bone healing as evaluated by radiography, densitometry, and biomechanical testing, in a canine radial gap defect model. In conclusion, the most extensive experimental work for veterinary patients has been performed in the dog and confirms the efficacy of BMP2 and OP-1 in augmenting bone formation.

1.6 Molecular interactions relevant to musculoskeletal and connective tissue

development and healing

The studies described above, performed in clinically relevant large animal models, provide a wealth of data demonstrating the safety and feasibility of the use of recombinant growth factors in the healing of defects in musculoskeletal connective tissues. Results of these studies confirm that healing of bone is augmented by the use of growth factors, and specifically, that bone healing in the presence of BMP2 and OP-1 appears to occur through the mechanism of endochondral ossification. However, the assessment of the potential for the use of these and other growth factors for novel clinical applications in the future lies in the ability of investigators to confirm the commonly asserted hypothesis that fracture healing in adults recapitulates the process of embryonic limb formation [40,89-95]. The potential clinical implications of the above-described studies lie in the fact that in both embryonic limbs and in fracture sites in adults, new bone is formed. A thorough understanding of the events occurring in the development of the embryonic limb may elucidate molecular targets for therapeutic intervention in the healing of fractures in adults. In addition, documentation of the molecular signaling

events occurring in experimentally created fractures has provided a wealth of information with the potential for clinical application [40,92,95].

By studying gene expression during healing of unstable diaphyseal fractures in mice,

Le et al [95] documented the sequence of molecular events associated with bone healing in both stabilized and unnstabilized fractures. These findings are significant in that the mechanisms of bone healing in stabilized and unstabilized fractures have been elucidated and clearly documented. Specifically, these studies document the role of BMPs in fracture healing by defining the signaling network of gene expression involved in their regulation. The discovery of the mechanisms of fracture healing mediated by BMPs provide a potentially powerful tool for therapeutic interventions in fracture repair in all species.

Interrelationships between BMP gene expression and embryonic skeletal morphogenesis have been well studied (Figure 1.3), and may provide a reference from which to consider the effects of particular BMPs on chondrogenesis and osteogenesis under various biological conditions, such as fractures.

The elucidation and repeated confirmation of these signaling mechanisms in embryonic tissues is essential to the understanding of repair, regeneration, and remodeling of adult tissues, and as such, provides the foundation for investigations into the potential clinical applications of the aforementioned growth factors in the healing of musculoskeletal connective tissues in veterinary medicine. The elucidation of the signaling interactions involving these growth factors and their targets may also identify possible targets for therapeutic modalities which are currently on the threshold of clinical

application, such as delivery of the genes encoding these proteins to injured tissues to augment healing through a biologically oriented approach.

1.7 Perspectives on the clinical applications of growth factors, disease-modifying agents and biologics

While introducing the concept of biologic approaches to tissue healing, the sequencing and cloning of a number of growth factors in the 1970’s and 1980’s also initially led to speculation about the potential feasibility of such applications, particularly across species. The fact that the amino acid sequences of the BMPs are highly conserved accounts for the lack of a clinically significant immune response associated with the use of rhBMPs, as well as the delivery of human BMP genes to cells in multiple species. The proteins of the BMP family are believed to have been conserved for approximately 600 million years [96,97]. The mRNA sequence for human BMP2 has 90% homology with the lapine BMP2 mRNA sequence, 89% homology with the equine sequence, 88% homology with the ovine sequence, 86% with the sequences of both the mouse and the rat, and 81% homology with the canine BMP4 gene sequence. The canine BMP2 gene sequence is not available on public genomic databases at this time. Consideration of these sequence homologies facilitates an understanding of the potential applications for recombinant growth factors across species. This therapy may hold particular promise for treatment of fractures and other musculoskeletal injuries, as the time during which therapeutic levels of the appropriate growth factor(s) would be required are considerably shorter than those required for chronic disease states.

periarticular region PTHrP BMPR-IA

perichondrium (-) PTC

PTH/H PTHrP BMP Receptor

proliferating chondrocytes S GLI m IHH a BMPR-IA prehypertrophic d R BMP chondrocytes 4 u n x hypertrophic Collagen Type X 2 chondrocytes Smad1/Smad 5

degenerative chondrocytes Modified from Vortkamp et al 1996, Zou et al 1997, and LeBoy et al 2001

Figure 1.3: Interactions between bone morphogenetic proteins (BMPs), Indian hedgehog protein (Ihh), parathyroid hormone-related peptide (PTHrP), parathyroid hormone-related peptide receptor, Smad intracellular transcription factors, and the transcription factor Runx2 in regulation of chondrocyte maturation in the growth plate.

1.8 Clinical implications for new therapies and devices

Despite the apparent successes described above in controlled research settings, in

2001 the orthopedic community was reminded that, “It has been almost 40 years since

Marshall Urist’s seminal observations on the bone induction principle were first published, yet we have not seen the common use of bone morphogenetic proteins (BMPs) in our patients.” [98] While there have been a number of challenges to be overcome in

the transition from preclinical studies in experimental animals to clinical trials in humans,

the first reports of human clinical trials are now published, reporting the use of

recombinant BMPs in treatment of open tibial fractures [99,100], tibial nonunions [101],

and spinal fusion [102-105].

Despite recent approval by the FDA and completion of clinical trials in human

patients, certain concerns remain associated with the development of these products for

more widespread clinical applications. Concerns exist surrounding optimization of the

dose of recombinant BMPs necessary to induce osteogenesis in humans, when compared

with that necessary to accomplish the same in rodents. However, preclinical studies have

shown that this is likely much less of a concern when using rhBMPs in dogs [106] (Table

1.2). These data demonstrate that there are differences in the effects of the same

recombinant BMP across species. For example, a smaller relative dose in dogs is

associated with shorter healing times than those in humans and non-human primates,

while the smallest relative doses, when administered to rats and rabbits, result in the

shortest healing times of all species studied thus far [106].

Another consideration in the further development of routine clinical use of

recombinant BMPs is the lack of consensus on an ideal carrier. Lieberman et al [35] have proposed that, “It is still unclear as to whether any of the currently known carriers have been truly optimized for clinical applications.” In order to achieve local effects of rhBMPs, the soluble recombinant protein must be immobilized at the site by being bound

to a carrier [107]. Of the numerous carriers investigated thus far, the absorbable collagen sponge (ACS) appears to hold the most promise for maximizing sustained release of rhBMP2 [108].

Species BMP Dose (mg/ml) Time to Healing in Presence of BMP Rat 0.02 2-3 weeks

Rabbit 0.02 3-4 weeks

Dog 0.04 6-8 weeks

Monkey 0.75-1.5 3-5 months

Human 1.50 4-6 months

Table 1.2: Dose of rhBMP and rate of bone healing in various species in the presence of recombinant bone morphogenetic proteins (BMPs) [109]. A dramatic difference is seen between healing times of rats and rabbits when compared with those of dogs. In addition, healing times in primates (both human and nonhuman) are longer than those of dogs.

The need for implantation of the ACS has led to the development of injectable carriers for rhBMPs, to ultimately allow for minimally invasive approaches to the surgical management of closed fractures. One of the most promising of these is an

 apatitic calcium phosphate cement paste known as α-BSM (ETEX Corporation,

Cambridge, MA). This product is approved in the United States for use in periodontal and craniofacial surgery, as well as in Canada and Europe for applications in orthopedic

surgery. This product is currently in Phase III clinical trials for the treatment of tibial

fractures in human patients in the US. Convincing data to support the clinical use of this

product were recently obtained using a rabbit ulnar osteotomy model [109]. In this study,

 osteotomies treated with rhBMP2 in an α-BSM carrier resulted in complete bony bridging 30 days postoperatively (a result not achieved in contralateral control limbs

 receiving buffer and α-BSM ). Biomechanical properties did not differ from those of

unoperated controls.

1.9 Clinically available therapies and devices

Despite the above concerns, in October 2001, rhOP-1 was approved by the Food

and Drug Administration (FDA) for use in the management of nonunion long bone

fractures in humans (OP-1 Implant, Stryker Biotech, Hopkinton, MA, http://www.op1.com; Table 1.3).The OP-1 Implant product is delivered in a collagen

carrier. It is packaged as a powder and reconstituted with sterile saline intraoperatively to

form a moist paste. The use of rh-OP1 in a collagen carrier has been shown to be a safe

and effective mechanism of delivery of this protein in segmental bone defects in dogs in a

number of preclinical studies [67-71]. These studies provide ample evidence to justify the

potential extra-label use of rh-OP1 for the management of delayed union and nonunion

fractures in canine patients in clinical settings.

Growth Potential Effect/FDA- Smallest Available Trade Name Cost Factor/Protein Approved Indications Form (Manufacturer) rhBMP2∗ Osteoinduction for Recombinant human InductOS Not (dibotermin management of protein on absorbable (Wyeth available alfa) nonunions in long collagen sponge carrier Pharmaceuticals) bones rhBMP2∗ Osteoinduction for 4.2 mg rhBMP2 in InFUSE Bone Graft/AC augmentation of lumbar collagen matrix (two 1” Graft/LT- S♣: ♣ spinal fusion x 2” ACS ) makes 2.8 CAGE $3500 ml graft (1.5 mg/ml (Medtronic Fusion rhBMP2) with tapered Sofamor Danek) cage: fusion cage (12 x 20 $3530 mm) rhBMP7∞ Induction of 2.5 mg rhOP-1 + 1 g OP-1 Implant $5000 + (rhOP-1#) osteogenesis in bovine type I collagen (Stryker Biotech) $200 recalcitrant nonunions (bone origin) shipping of long bones per unit rhPDGF-BB† Formation of 0.01% gel, 15 g tubes Regranex $540 (becaplermin) vascularized (Ortho-McNeil granulation tissue in Pharmaceutical) diabetic foot ulcers - ♠ CaSO4 Osteoconductive and 3 cc DBM in moldable Allomatrix DR $520 Augmented osteoinductive for carboxymethylcellulose (Wright Medical Demineralized osteogenesis in small, putty + option to add Technologies) Bone Matrix periarticular fractures species-specific (DBM) System cancellous bone chips +/- autologous bone marrow Injectable Induction of 20 cc DBM + system Ignite ICS $1250 Cellular osteogenesis in delayed for autologous marrow (Wright Medical Scaffold union and nonunion delivery for Technologies) fractures percutaneous administration

Table 1.3: Commercially available growth factors and their respective costs (as quoted for purchase on a unit basis).

In January 2002, the Orthopedic and Rehabilitation Devices Panel of the FDA approved the use of rhBMP2 in an ACS system with a tapered metal cage for use in lumbar interbody spinal fusion in humans (InFUSE Bone Graft + LT-CAGE Lumbar

Tapered Fusion Device, Medtronic Sofamor Danek, Warsaw, IN). Finally, in November

2002, the FDA recommended approval of rhBMP2 in an absorbable collagen sponge

(ACS) carrier (rhBMP2/ACS). The product will be marketed under the trade name

InductOS. Final FDA approval is pending.

In contrast to the lack of clinical data in the veterinary literature on the use of growth factors in surgery, a considerable amount has been published of late documenting the first human clinical trials using recombinant human BMPs. A prospective, randomized clinical trial comparing rhOP-1 with fresh bone autograft in 122 human patients with 124 nonunion tibial fractures reported 81% of the rh-OP1-treated nonunions and 85% of those receiving autogenous cancellous bone graft were judged to be successful by previously defined clinical criteria [101]. Since greater than 20% of patients treated with bone autografts had chronic donor site pain the use of recombinant protein offered morbidity advantages. In a prospective, controlled, randomized study of

450 human patients in an international clinical trial, the use of rhBMP2/ACS in open tibial fractures resulted in significantly fewer hardware failures, fewer infections, and faster wound healing at six weeks postoperatively [101].

1.10 The clinical future for therapies and devices

While there is no rhBMP2/ACS product similar to InductOS available on the veterinary market, Phase II clinical trials are in progress at select institutions evaluating a formulation of rhBMP2 targeted for FDA approval for use in canine patients. This product is being developed by Fort Dodge Animal Health, a subsidiary of Wyeth

Pharmaceuticals. At this time, the product is not available to surgeons not participating in these clinical trials. No data has been published on the use of this product to date. Use of the human product by veterinarians is permissible, but the dose and pharmacokinetics of the protein release are likely to differ among species and product expense is often beyond our budgets.

Despite the promising results seen thus far with the use of recombinant proteins, their use still requires expensive manufacturing and has limitations associated with cross- species use, a very relevant restriction for veterinary medicine. In recent years, considerable effort has been directed toward the development of products to augment the ability of clinicians to use autologous sources to augment bone and soft tissue healing.

These endeavors have led to numerous advances in devices to facilitate this process.

Among the most prominent of these is the refinement of techniques for the creation of platelet concentrate, as well as commercially available systems for the selection and concentration of BMDMSC from autologous marrow samples. These autologous products can supplement osteoconductive scaffolds with endogenous osteoinductive elements and cells present in blood products and cancellous bone. The goal for the use of all of these products is to minimize or eliminate harvest site morbidity associated with the

harvest of massive volumes of cancellous bone in both veterinary and human patients.

Platelet concentrate is an autologous source of growth factors obtained from whole blood

via a specialized centrifugation process that selectively concentrates platelets. The

application of platelet concentrate to a wound, fracture site, or bone-implant interface

allows for a concentration of growth factors at the site of interest. Although results may

vary between species, concentrations of three to five times the peripheral blood platelet

count can be achieved. In horses, concentration of platelets correlated to concentration of

TGF-β2 and with further filter concentration can achieve a 14-fold increase in TGF-β2

concentration over whole blood [28]. These concentrates can promote healing. Following

degranulation at the site of deposition of platelet concentrate, platelets release PDGF,

TGF-β, and vascular endothelial growth factor (VEGF), among other growth factors.

When PDGF binds to endothelial cells, the process of capillary ingrowth at the site is triggered. Enhancement of osteogenesis results when TGF-β binds to osteoblasts and to

BMDMSC, stimulating mitosis and initiating the production of osteoid [110]. To our

benefit, a large number of the preclinical studies evaluating the use of platelet concentrate

for human patients have been performed in animal models [111-113].These experimental

studies, performed in dogs [112,113] and in miniature pigs [111], demonstrated superior

healing at bone-titanium interfaces around bone implants [112,113] and enhanced healing

of bone graft sites [111]. In addition, the use of platelet concentrate gel for the

augmentation of healing of chronic wounds in the distal limb of a horse has recently been

reported [114]. Commercially available systems for the preparation of platelet

concentrate for clinical use include the Secquire® Cell Separator (PPAI Medical, Fort

Myers, FL), the Cell Saver 5® System (Haemonetics Corp., Braintree, MA); the

SmartPReP Platelet Concentrate system (Harvest Technologies, Plymouth, MA); and the

Gravitational Platelet Separation (GPS) System (Biomet Merck, Dordrecht, The

  Netherlands). The Secquire and Cell Saver 5 systems have successfully concentrated

platelets from equine whole blood using human protocols [28].

Another promising application of autologous sources for the healing of fractures

lies in the use of autologous BMDMSC. These are pluripotential cells residing in both

juvenile and adult bone marrow. Under the influence of growth factors, mainly the BMPs

and TGF-β, these cells can differentiate into osteoblasts, chondrocytes, adipocytes, and

myoblasts [115].These cells can be delivered to individual patients via autologous bone

marrow injection into fracture sites for the augmentation of healing of delayed union or

nonunion fractures [35,115,116]. It has been demonstrated that BMDMSC will home to

fracture sites when administered systemically [117]. Accelerated healing of diaphyseal

defects in canine tibiae has been demonstrated following autologous bone marrow

injection [118-120].

Recently, the Orthopaedic Research Center at the Cleveland Clinic Foundation

and the Center for Stem Cell and Regenerative Medicine in Cleveland, OH have

announced the availability of the Cellect system, marketed by DePuy Inc. This system

allows for the collection of bone marrow by aspiration and subsequent selection for and

concentration of BMDMSC patient-side for controlled delivery to fracture sites. This

specific technology, called Selective Cell Retention (SCR) Technology has been shown

to heal critical-sized segmental defects in canine femora as effectively as autograft bone

[121,122]. In this study, SCR technology was used successfully with canine bone marrow

and canine bone allografts. Its use in veterinary medicine seems a logical extension.

These systems (InFUSE Bone Graft/LT-CAGE, the systems described here for the creation of platelet concentrate, and the Cellect system and SCR technology for concentration of BMDMSC) are considered devices, rather than drugs, by the FDA, by virtue of the fact that they utilize autologous sources to augment healing. This is significant in that the number of these products available to veterinary clinicians will likely increase at a faster rate than will the number of new drugs, as devices characteristically have a faster progression through the process of FDA approval than do pharmaceuticals. This is also representative of the trend initiated in the 1970’s with the first cloning and production of recombinant growth factors, which continues today with the above-described trends toward the continued search for improved, autologous sources of therapy to modulate healing.

Finally, a discussion of therapies with potential clinical applications in musculoskeletal and connective tissue healing would not be complete without a discussion of gene therapy. Gene therapy can be used to deliver the gene encoding the recombinant proteins mentioned here and subsequently use the patient’s cells to regenerate the growth factors. An application of gene therapy which may be the closest to clinical use in veterinary patients is the delivery of the gene encoding IL-1Ra for the treatment of osteoarthritis and the use of IGF-1 for augmentation of cartilage healing.

Mitigation of clinical signs, as well as changes in more objective laboratory parameters, were recently demonstrated with adenovirus-mediated delivery of the gene encoding equine IL-1Ra in an experimental equine in vivo model of osteoarthritis [123]. The potential advantage to the delivery of the gene encoding IL-1Ra, as opposed to treatment of patients with the recombinant human protein, is the short half life of this protein in vivo [124] and the lack of oral availability of this formulation, which requires that it be administered via daily subcutaneous injections [125]. In addition, a recently published in vitro study demonstrated sustained or increased protection from IL-1 in cells genetically modified to express IL-1Ra, when compared with human synovial fibroblast cultures incubated with recombinant human IL-1Ra in varying doses. Because of the similarities between the clinical signs and laboratory parameters of osteoarthritis and other inflammatory arthritidies in human and veterinary patients, it is reasonable to consider that IL-1Ra therapy, either by recombinant protein administration, or by gene delivery, may be available to veterinarians for clinical applications in the future.

1.11 Conclusions

Approval by the FDA for the use of growth factors in surgery has occurred largely in response to data from human clinical trials, and from preclinical studies, most of which were performed in dogs with excellent results, as described above. A select group of commercially available products containing recombinant growth factors and their active ingredients, in addition to their costs (for purchase of a single unit) is presented in Table

1.3. The availability of recombinant growth factors on the veterinary market may be on

the horizon. However, careful consideration of data published to date using these products in clinically applicable animal models and in human patients is warranted, as these comprise the body of our current knowledge about this diverse group of proteins.

While extrapolations from human clinical studies to veterinary medicine are difficult and imprecise, a prudent approach for clinicians investigating progressive treatment options utilizing a biologic approach to the healing of musculoskeletal and other connective tissues may find a wealth of clinically applicable data in the scientific literature preceding human clinical trials of the growth factors discussed above. It is anticipated that these vast basic science and clinical databases will set the stage for the judicious clinical application of these products in veterinary patients in the near future.

CHAPTER 2

GENE-MEDIATED OSTEOGENIC DIFFERENTIATION OF STEM CELLS BY

BMP2 OR BMP6

2.1 Summary

Bone marrow-derived mesenchymal stem cells (BMDMSC) hold promise for targeted osteogenic differentiation and can be augmented by delivery of genes encoding bone morphogenetic proteins (BMP). The feasibility of promoting osteogenic differentiation of BMDMSC was investigated using 2 BMP genes in monolayer and 3- dimensional alginate culture systems. Cultured BMDMSC were transduced with E1- deleted adenoviral vectors containing either human BMP2 or BMP6 coding sequence under cytomegalovirus (CMV) promoter control (17:1 multiplicities of infection (MOI)) and either sustained in monolayer or suspended in 1 ml 1.2% alginate beads for 22 days.

Adenovirus (Ad)-BMP-2 and Ad-BMP-6 transduction resulted in abundant BMP-2 and

BMP-6 mRNA and protein expression in monolayer culture and BMP-2 protein expression in alginate cultures. Ad-BMP-2 and Ad-BMP-6 transduced BMDMSC in monolayer had earlier and robust alkaline phosphatase-positive staining and

mineralization and were sustained for a longer duration with better morphology scores than untransduced or Ad-ß-galactosidase-transduced cells. Ad-BMP-2- and to a lesser degree Ad-BMP-6-transduced BMDMSC suspended in alginate demonstrated greater mineralization than untransduced cells. Gene expression studies at day 2 confirmed an inflammatory response to the gene delivery process with up-regulation of interleukin 8 and CXCL2. Up-regulation of genes consistent with response to BMP exposure and osteogenic differentiation, specifically endochondral ossification and extracellular matrix proteins, occurred in BMP-transduced cells. These data support that transduction of

BMDMSC with Ad-BMP-2 or Ad-BMP-6 can accelerate osteogenic differentiation and mineralization of stem cells in culture, including in three-dimensional culture. BMP-2- transduced stem cells suspended in alginate culture may be a practical carrier system to support bone formation in vivo. BMP-6 induced a less robust cellular response than

BMP-2, particularly in alginate culture.

2.2 Introduction

The capacity of bone marrow-derived mesenchymal stem cells (BMDMSC) to differentiate into osteoblasts, as well as into cells of other mesenchymal lineages, has been well documented [16,17,27,53,74,78,93,95]. Delivery of recombinant bone morphogenetic protein-2 (BMP-2) has induced expression of bone markers [45,57,90] and mineralization [45] in monolayers of bone marrow stromal cells cultured in differentiating medium containing dexamethasone [45,90] and/or β-glycerophosphate and ascorbic acid. [45,57,90] Gene delivery of BMP-2 has augmented the expression of

alkaline phosphatase in monolayers of bone marrow-derived mouse mesenchymal cell

lines, particularly with ascorbic acid media supplementation [66]. Gene delivery studies

performed with human bone marrow stromal cells in monolayer culture revealed that

transduction with Ad-BMP-2 prior to deliberate osteogenic induction in vitro markedly

decreased BMP-2 protein production when compared with cells induced to osteogenic

differentiation prior to adenoviral transduction [81]. Additionally, gene delivery of BMP-

2 to human BMDMSC induced minimal alkaline phosphatase activity in monolayers

unless supplemented with a differentiating medium of dexamethasone and ascorbic acid

[81]. A few studies have demonstrated the potential for BMDMSC genetically engineered

to produce BMP-2 to induce bone formation in vivo through the process of endochondral

ossification when placed in muscle [32,60,66,76,79] and bone [21,22,40,60,61,76]

defects. The influence of BMP-6 on osteogenic differentiation and the biology of

BMDMSC have not been studied in vitro. In athymic nude rats, intramuscular injection

of Ad-BMP-6 induced bone formation that resembled intramembranous ossification [54]

and Ad-BMP-6 gene delivery accelerated bone formation in a rabbit ulna osteotomy

model [14]. These studies suggest a role for BMP-6 in osteogenesis, possibly through

osteogenic differentiation of immature repair cells. In vitro, recombinant BMP-6 has been shown to play a definitive role in mineralization of growth plate chondrocytes [2,16,88].

Three-dimensional systems may facilitate placement and retention of cells at desired delivery sites. Transfer of BMDMSC can augment bone formation in vivo

[21,22,40,60,61,76] using three-dimensional systems such as collagen gels [21,32] and

demineralized bone matrix [60,61]. Alginate gels have been used to suspend BMDMSC

[22,31,91] but osteogenic differentiation by BMPs in vitro in this system, as well as the ability of this system to support BMDMSC cell health in vitro, has not been evaluated.

Alginate has been used effectively to deliver human, murine, and lapine BMDMSC to heal musculoskeletal connective tissues, particularly for the induction of chondrogenic differentiation [31,57,68-70,92,97]. In these studies, cells had been preconditioned via exposure to one or more factors previously demonstrated to induce osteogenic and/or chondrogenic differentiation, e.g. dexamethasone, ascorbate-2-phosphate, BMP-2, BMP-

9, TGF-β1, and/or TGF-β3. The ability of an alginate system to support transduced stem cells and accelerate osteogenic differentiation in suspended, isolated cells in vitro has not been reported. Adenoviral vectors were selected to deliver the genes in this study because they are efficient, safe mediators of gene transfer to both dividing and non-dividing cells and have been used effectively to deliver BMP-2 to BMDMSC

[15,21,22,40,60,61,66,76,79].

Our study focuses on the ability of two BMP genes, BMP-2 and BMP-6, to accelerate osteogenic differentiation of BMDMSC which have not been treated with osteogenic or chondrogenic differentiating agents, in monolayer and suspended in a 3- dimensional alginate culture system. We investigated the potency of adenovirus transduction of stem cells with BMP-2 and BMP-6, with the hypothesis that both BMP-2 and BMP-6 could successfully promote early osteogenic differentiation of untreated cells from immunocompetent animals, including in a 3-dimensional alginate culture system that could be readily implanted in vivo.

2.3 Methods

Study design

Bone marrow-derived mesenchymal stem cells were isolated and expanded in monolayer or placed in 1.2% alginate three-dimensional cultures. Cells, in triplicate and in both culture systems, were evaluated at four time points (immediately prior to adenoviral transduction (day 0) and on days 2, 12, and 22 after seeding) for 5 experimental groups; 1) untreated controls (No Ad); 2) adenovirus-reporter gene construct encoding bacterial β-galactosidase to document efficiency of adenoviral transduction (Ad-LacZ); 3) adenovirus-reporter gene construct carrying the gene encoding recombinant firefly luciferase to quantify the intensity and duration of gene expression (Ad-Luc), 4) Ad-BMP-2 vector and 5) AdBMP-6 vector. Outcome assessments included transduction efficiency, quantification and duration of gene expression, and assessments of osteogenic differentiation using cytomorphology, cytochemistry, and gene expression analysis.

Generation of adenoviral vector constructs

Recombinant adenoviral vectors containing either a 1547 base-pair open reading frame segment of human BMP-2 or a 1539 base-pair open reading frame segment of human BMP-6 [20], under the control of the cytomegalovirus promoter, were generated

[14]. Expression of transgenes was verified in cell culture.

Cell culture systems

Mesenchymal stem cells from 5 adult horses (aged 5-9 years) were obtained from

bone marrow aspirates and expanded as primary cells in monolayer culture. The

pluripotentiality of these BMDMSC was confirmed by culture in controlled osteogenic,

chondrogenic, and adipogenic media cocktails containing dexamethasone with ascorbate,

rhTGF-β1, and dexamethasone with insulin and indomethacin, respectively.28,76,82,100,a

Cells were seeded into 24-well plates at 2.2 x 104 cells per well and maintained in

Dulbecco’s Modified Eagle Medium (DMEM; Gibco) supplemented with 10% fetal

bovine serum, sodium penicillin at a concentration of 50 units/ml, streptomycin at a

concentration of 100 units/ml, and L-glutamine at a concentration of 29.2 mg/ml

(supplemented DMEM).

At 75% confluence, wells were assigned, in triplicate, to monolayer or alginate

systems and one of the five groups. Adenoviral transduction of all wells was performed

in monolayer at a multiplicity of infection (MOI) of 17:1b at 37°C for a transduction time

of two hours, washed and allowed to incubate overnight to achieve expression of

transgene. Cells assigned to alginate groups were suspended in a 1.2% solution of sodium

alginate (Sigma® Alginic Acid Sodium Salt, P/N A-0682, Sigma Chemical Company, St.

Louis, MO) at a concentration of 2.2 x 104 cells/ml. Aliquots of the cell-alginate

suspension in 1 ml volume were placed into a 102 mM solution of CaCl2 for 10 minutes

to solidify and were rinsed with 0.9% NaCl [70].

Transduction efficiency

Transduction efficiency was evaluated by histologic quantitative assessment of monolayer cell culture systems in each of five microscopic fields (X200) by using 5- bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) staining in the Ad-LacZ and

No Ad group. Transduction efficiency (%) of cells in monolayer and alginate cultures was calculated or estimated, respectively, as percent of cells with X-gal uptake were divided by the total number of cells per field.

Transgene and protein expression

Luciferase production in the Ad-Luc group was quantified using an in vivo

 imaging system (IVIS , Xenogen Corporation, Alameda, CA) at 2 hours post- transduction (day 0), and on evaluation days 2, 12, and 22. Luciferase production was quantified as flux (photons of light produced per second per square centimeter per steradian, photons/sec/cm2/sr) [65].

Transgene expression of human bone morphogenetic protein 2 (hbmp2) and human bone morphogenetic protein 6 (hbmp6) were quantified using real-time reverse- transcription polymerase chain reaction with the ABI PRISM 7000 Sequence Detection

System (Applied Biosystems, Foster City, CA).1 Total RNA from cells was isolated,

 using a commercially available kit (RNEasy , QIAGEN Inc., Valencia, CA), from groups 1, 4, and 5 in the monolayer cultures at days 0, 2, 12, and 22. Mean fold change in hbmp2 and hbmp6 gene expression at days 2, 12, and 22 post-transduction of cells was

calculated, relative to expression at day 0 and relative to expression of the endogenous

-∆∆C control gene encoding eukaryotic 18SrRNA, using the 2 T method [64]. The primer and probe sets were designed by Applied Biosystems (Foster City, CA) as part of the

 TaqMan Gene Expression Assays.

Aliquots of media from both cells in monolayer and alginate cultures were frozen

at -80°C on days 0, 2, 12, and 22. Production of BMP-2 and BMP-6 proteins were

quantified using enzyme-linked immunosorbent assays (ELISAs) for recombinant human

  (rh) BMP-2 (Quantikine , R&D Systems, Minneapolis, MN) and rhBMP-6 (DuoSet ,

R&D) and expressed as pg/ml/day.

Cytomorphometry

Cell morphology scoring was performed on monolayer and alginate cultures on days 0, 2, 12, and 22 to assess cell health. Three representative fields (200X magnification) were assigned 0 (markedly elongated cell), 1 (moderately elongated cell, fibroblastic), 2 (equine BMDMSC with normal appearance), 3 (rounded, beginning to detach), 4 (dead, detached) in monolayer cultures and 0 (round, smooth, bright cell membrane), 1 (cell membrane less bright and smooth), 2 (round cell with dull appearance to cell membrane), 3 (slightly crenated cell with dark cell membrane), 4 (markedly crenated cell with dark cell membrane) in alginate cultures.

Cytochemistry

On days 0, 2, 12, and 22, triplicate monolayer wells were fixed and stained for alkaline phosphatase activity (Sigma Kit No. 85, Sigma Chemical Company, St. Louis,

MO; AP) and mineral (von Kossa method; VK). Digital images, immediately before and after staining, were scored for stain uptake (three fields at 200X magnification) according to the following grading scheme: 0 = no stain uptake other than AP background staining, which is a characteristic of wild-type BMDMSC:81 1 = 1- 25%; 2 = 26-50%; or 3 = 51-

100% of cells in the microscopic field of view demonstrating stain uptake). Alginate constructs were assessed for von Kossa staining as follows: 0 = no staining, 1 = mild staining, 2 = moderate staining, 3 = marked staining on days 0, 2, 12, and 22.

BMDMSC gene expression analysis

 Total RNA was isolated (RNeasy , QIAGEN) from BMDMSC from Ad-BMP-2,

Ad-BMP-6, Ad-Luc, and untreated (No Ad control) monolayers on days 0, 2, and 12 for

 gene expression by an equine-specific microarray (Custom Equine GeneChip ,

Affymetrix, Inc., Santa Clara, CA) and prepared for application to microarrays using established methods [41]. Each equine microarray represented 3,098 separate genes, of which 1945 genes were annotated by Basic Local Alignment Search Tool (BLAST) and functionally classified (http://www.ncbi.nlm.nih.gov/BLAST/) [44].

Statistical Analysis

Objective values (transduction efficiency, BMP concentrations, and flux) were compared among groups using a multiple two-tailed repeated measures analysis of covariance (MANCOVA) with Tukey’s post hoc comparisons (Statistica, Stat Soft,

Inc., Tulsa, OK). Scored data (cytomorphology, alginate X-gal staining, alkaline phosphatase staining and von Kossa staining) were compared using the Mann-Whitney test U test (Statistica). For all statistical analyses, p values of <0.05 were considered significant.

2.4 Results

Confirmation of Gene Transduction: LacZ

Transduction efficiency (β-galactosidase-positive cells) was greatest in monolayer cells, with a mean of 91.5% ± (8.3%) at day 2 and a mean of 69.6 % ± (6.7%) at day 12.

Monolayer cells transduced with Ad and without BMP growth factor support (Ad-LacZ) died after day 15. Transduction efficiency was approximately 25% for Ad-LacZ cells in alginate on days 2, 12 and 22 (Figure 2.1). The median X-gal score for No Ad controls was 0 at all time points, indicating no staining.

Transgene and Protein Expression

Gene transduction, gene transcription, and translation were confirmed for Ad-

BMP2- and Ad-BMP6-transduced BMDMSC in monolayer culture. Maximum BMP-2

concentration was detected at day 22 (32,408 ± 23,501 [SEM] pg/ml) and maximum

BMP-6 concentration (47,129 ± 26,275 [SEM] pg/ml) was detected at day 2. Increases were significantly greater than baseline and No Ad controls at all time points (p < 0.02;

Table 2.1).

For cells in alginate constructs, transduction of cells and sustained BMP2 production in the AdBMP-2 group was confirmed by an increase in BMP-2 protein concentration over No Ad (705 pg/ml vs. 79 [SEM] pg/ml) at day 2, at day 12 (53,917 pg/ml vs. 117 [SEM] pg/ml), and at day 22 (36,666 pg/ml vs. 0 pg/ml). Increases were significant at all time points (p < 0.03; Table 2.1; Figure 2.2). For cells in alginate constructs, BMP6 protein production was below detection in media from Ad-BMP-2- and

Ad-BMP-6-transduced cells.

Figure 2.1: Appearance of equine BMDMSC in monolayer (upper two rows) and alginate (lower two rows) culture. Mono = monolayer cultures; ALG = alginate cultures; N/A = not available: transduced cells (β- galactosidase-positive) in AdLacZ group are lifting off of the plate at day 12 and did not survive to day 22.

Variable Culture Group Day System 0 2 12 22 BMP-2 expression Monolayer No Ad 303.7 ± 12.5 ± 4.7 344.4 ± 107.8 ± (pg/ml) 91.5 130.0 48.1 Ad- 0 ± 0 7054.1 ± 8198.1 ± 57,564.3 BMP-2 2661.9 2732.7 ± 23,495.6 Ad- 0 ± 0 425.0 ± 300.5 ± 894.2 ± BMP-6 188.9 173.7 337.4 Alginate No Ad 217.3 ± 59.5 ± 14.9 181.3 ± 90.7 0 ± 0 88.7 Ad- 217.3 ± 675.2 ± 53,917.1 ± 36,665.7 BMP-2 88.7 337.6 31165.9 ± 21,194.0 Ad- 5.0±10.0 518.9±497.4 580.4±663.1 8.1±14.9 BMP-6 BMP-6 expression Monolayer No Ad 0 ± 0 24.9 ± 24.9 114.5 ± 14.2 ± (pg/ml) 114.2 14.2 Ad- 2.8 ± 2.8 0 ± 0 223.7 ± 0 ± 0 BMP-2 224.0 Ad- 0 ± 0 44,332.9 ± 1356.7 ± 15,229.9 BMP-6 19,791.5 553.7 ± 7614.9 Alginate No Ad ND ND ND ND Ad- ND ND 85.5±148.0 ND BMP-2 Ad- ND ND ND ND BMP-6 Luciferase Monolayer No Ad 8.1 x 104 5.5 x 106 ± 3.5 x 106 ± 2.8 x 106 expression [flux ± 3.9 x 106 2.5 x 106 ± (photons/sec/cm2/sr)] 5.8 x 104 2.0 x 106 Ad-Luc 1.5 x 106 1.4 x 108 ± 8.6 x 107 ± 9.6 x 107 ± 9.7 x 107 6.1 x 107 ± 1.0 x 106 6.8 x 107 Alginate No Ad 2.3 x 104 1.4 x 104 ± 2.3 x 104 ± 1.9 x 104 ± 1.8 x 103 8.3 x 103 ± 8.3 x 103 6.7 x 103 Ad-Luc 2.3 x 105 2.7 x 104 ± 1.7 x 104 ± 1.7 x 104 ± 9.4 x 103 6.2 x 103 ± 8.1 x 104 6.2 x 103 ND = not detected

Table 2.1: Transgene protein expression (mean ± SEM) of bone marrow-derived mesenchymal stem cells.

AdLuc Untreated

Alginate construct, day 1

Monolayer, day 2

Figure 2.2: Detection of luciferase in equine BMDMSC in monolayer cultures and alginate constructs.

Luciferase production by Ad-Luc-transduced cells in monolayer cultures was

greater than No Ad control cells on days 0, 12, and 22 (p < 0.05; Table 2.1; Figure 2.2).

Luciferase detection in cells in alginate constructs was ~ 10-fold lower than that in

monolayer systems and was greater than that in No Ad controls at days 1 (p < 0.000004)

and 2 (p < 0.005, Table 2.1). Gene transduction, gene transcription, and translation were

detected as early as two hours post-transduction in both systems (Figure 2.2). Gene

expression of hbmp2 and hbmp6 was increased at all time points for cells in monolayer culture as compared to baseline expression at day 0 (Figure 2.3). Human bmp2 gene expression was 2514-, 20-, and 4-fold increased at days 2, 12, and 22, respectively.

Human bmp6 gene expression was 6236-, 1780-, and 644-fold increased at days 2, 12, and 22, respectively.

Osteogenic differentiation – Histochemistry and assessment of cytomorphometric analysis

Monolayer cells transduced with Ad-BMP-2 and Ad-BMP-6 had significantly greater alkaline phosphatase scores at days 12 and 22 and von Kossa staining at day 12 than untreated cells. Cells transduced with Ad-BMP-2 and Ad-BMP-6 had significantly greater alkaline phosphatase and von Kossa staining on day 12 than Ad-LacZ cells at day

12. However, comparisons at day 22 were not possible, as cells in Ad-LacZ monolayer cultures died prior to day 22.

For cells in alginate constructs, von Kossa staining was lower than monolayer cultures. The degree of von Kossa staining increased in all groups by day 22 and staining occurred earlier in Ad-treated groups. In addition, von Kossa staining appeared greater in the Ad-BMP-2 alginate constructs, and to a lesser extent than the Ad-BMP-6 alginate constructs, when compared with the No Ad constructs (Figure 2.4B). In general, early staining of mineral appeared at the edges of the von Kossa-stained alginate constructs.

This was seen as early as day 12 and to a greater degree in Ad-BMP-2 alginate constructs and to lesser and varying degrees at days 22 in No Ad, Ad-LacZ, and Ad-BMP-6 alginate constructs. Varying degrees of punctate staining were seen in other areas of the constructs, to a greater degree in Ad-BMP-2 and Ad-BMP-6 alginate cultures (Figure

4B). This was different from the staining seen in monolayer cultures, in which alkaline

phosphatase staining was more diffuse in Ad-BMP-2 and Ad-BMP-6 cultures at days 12 and 22, and von Kossa staining was localized to prominent mineralized bodies at the same time points in Ad-BMP-2 and Ad-BMP-6 cultures (Figure 2.4A).

d d d BMP d 22 A Marker d d0 d2 No A d12 No A d22 No A d2 AdBMP6 d12 AdBMP6 hbmp6 300 bp expression 150 bp 50 bp 18SrRNA expression

Figure 2.3: Representative agarose gel (1.2%) electrophoresis of real-time RT-PCR products demonstrating increased intensity of bands representing cells transduced with Ad-BMP-2 (top panel) and Ad-BMP-6 (middle panel) when compared with baseline expression in untreated (No Ad) BMDMSC and similar expression of 18SrRNA in all groups (bottom panel).

Day 2 Day 12 Day 22 No Ad Ad BMP2 Ad BMP6

Figure 2.4: A. Comparison of von Kossa staining of alginate cultures. B. Comparison of monolayer cultures stained with alkaline phosphatase and von Kossa. The day 22 AdBMP6 culture is an example of an unstained culture. Mono = monolayer cultures; ALG = alginate cultures; N/A = not available: cells in AdLacZ group did not survive to day 22.

Overall, at the beginning of the study (day 0), all cells, in both monolayer and alginate cultures, appeared healthy. There was a gradation at days 2, 12, and 22 to a more senescent appearance of the cells, with a progression toward an increased number of detached cells over time. At day 22, there was significantly poorer morphology (p < 0.05) in the Ad-BMP-6 (median score: 3) and Ad-LacZ groups (median score: 4), when compared with the Ad-BMP-2 group (median score: 2).

Equine BMDMSC gene expression analysis

A mean of 55% of the genes on the array were expressed in BMDMSCs regardless of treatment, exhibiting a strong mean signal intensity of 3,494 ± 35 (standard error of the mean) and indicating consistency of array function and processing. The number of sequences detected was similar among groups for the number of expressed, nonexpressed, and marginally expressed genes.

Genes that were up- or down-regulated among groups and days are listed in

Appendix A (Table A1). The inflammatory genes, such as interleukin 8 (IL8) and

CXCL2, were up-regulated early on day 2 in cells transduced with adenovirus, including

AdLuc. Genes associated with a response to high exogenous BMP exposure, such as

Smad6, Cdc42 guanine nucleotide exchange factor (GEF) 9, BMP6 precursor, ALK5 for

TGF beta receptor type I, inhibin beta A subunit, and fibulin 1 (FBLN1, transcript variant

C) were up-regulated in AdBMP groups as early as day 2. Genes associated with cell differentiation and osteogenic differentiation, such as cartilage oligomeric protein

(COMP), procollagen alpha 1 (I), inhibin beta A subunit, BMP6 precursor, Smad6, procollagen alpha-1 type III precursor (COL3A1), parathyroid hormone-related peptide, keratinocyte growth factor (fgf-7), FBLN1, and retinoic acid receptor responder

(RARRES1), were up-regulated by day 12 in BMP-2 and BMP-6 groups. When considering all treatments and both time points, 35 sequences of interest in four function categories (endochondral osteogenesis, chondrogenesis, inflammation, and other) were up-regulated greater than 3-fold and 15 sequences of interest were down-regulated greater than 3-fold (Appendix A, Table A1). There were several sequences relevant to

musculoskeletal development, injury, disease, and repair whose gene expression was unchanged, such as aggrecan core protein, biglycan, decorin, dermatan sulfate proteoglycan II, fibronectin, insulin-like growth factor II, Sox9, and type II collagen

(Appendix A, Table A2).

2.5 Discussion

This study demonstrated the relative potential for induction of robust osteogenic differentiation of BMP-2 and BMP-6 delivered to unconditioned BMDMSC by an adenoviral vector in both monolayer and 3-dimensional alginate systems. In alginate suspension, osteogenic differentiation was less robust than monolayers manifested as greater mineralization and greater BMP-2 protein production in Ad-BMP-2 monolayer cultures as compared with Ad-BMP-2 alginate cultures. However, BMP-2 produced greater osteogenic differentiation of unpreconditioned BMDMSC than BMP-6. In particular, the results of this study demonstrated that BMP-2 expression sustains cell health and viability, and correlates with osteogenic differentiation, as adenoviral controls underwent senescence. This is supported by the work of others [50]. It has been proposed that the supportive effects of BMP-2 on osteoprogenitor cells may function through mechanisms similar to those demonstrated by BMP-2 in the development of sympathetic neurons [45]. In terms of potential in vivo applications, our findings suggest that cells genetically modified to express increased levels of BMPs should be delivered to desired anatomic sites before the cells begin to senesce. Our results suggest that the optimal time

for delivery of these cells is either at or after day 2, and prior to day 12. These findings demonstrate that BMDMSC are somewhat difficult to maintain under in vitro conditions.

Based on our data, we propose that early BMP2 expression is necessary and sufficient to induce osteogenic differentiation of these cells in culture. Greater expression at the later time points is likely the result of the autocrine and paracrine effects of the

BMP2 protein secreted by the cells in culture, which were originally bone marrow- derived mesenchymal stem cells (BMDMSC), but have now undergone differentiation into osteoblasts in response to the initial BMP2 production. Bone morphogenetic protein production may have been perpetuated in these cultures, resulting in the most dramatic increase in expression of BMP-2 at the latest time point evaluated. We further propose that the BMP-2 protein production produced as the result of bmp2 gene delivery in this study may have exceeded the threshold level of BMP necessary to induce osteogenic differentiation of BMDMSC, therefore the fluctuations in BMP-2 levels may have had little influence on outcome, i.e. osteogenic differentiation.

Our findings, in concert, support osteogenic differentiation of BMDMSC by

BMPs. Changes in gene expression across time points in this study revealed up- regulation of the genes encoding Smad6, BMP6 precursor, ALK5 for TGF beta receptor type I, inhibin (beta A subunit), fibulin 1 (transcript variant C), retinoic acid receptor responder 1 (RARRES1) and procollagen alpha 1 (I) (COL1A1) in our BMP-treated wells; these findings are consistent with osteogenic differentiation

[4,9,27,30,34,36,46,48,52,62,69,72,93,101]. (Please see Appendix A, Table A1.) The acceleration of functional development of BMDMSC into osteoblasts, manifested by

alkaline phosphatase expression and formation of mineralized bodies (also referred to as colony forming unit-osteoblast, CFU-O) [12] demonstrated classical osteogenic differentiation by BMPs in BMP-transduced cells in vitro. This is supported by the gene expression changes in our BMDMSC associated with BMP exposure, as demonstrated by the microarray data described above.

Transduction of BMDMSC using an adenoviral vector-reporter gene construct incited an early inflammatory response and later senescence of cells in monolayer and alginate cultures. This was supported by our gene expression studies, histology, and assessments of reporter gene products. We propose that transduction of cells with adenovirus appears to be physiologically stressful to cells based on their senescence in these culture systems when compared with cells in the untreated (No Ad) control and Ad-

BMP-2 groups, in light of the up-regulation of gene expression of the inflammatory mediators CXCL2 and IL-8. Expression of other genes associated with inflammation was up-regulated at later time points, and in BMP-2-treated groups as well. Some of this effect may be due to the vector inciting an inflammatory and/or apoptotic response, as evidenced by our gene expression studies, which showed up-regulation of the genes encoding p53-responsive gene 1 (PRG1) [50] nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor alpha (NFKBIA) [5] interferon regulatory factor 1

(IRF1) [82], CXCL2 [38] tumor necrosis factor alpha (TNF-α) [34,39], matrix metalloproteinase-3 (MMP-3) and interleukin-8 (IL-8), at early time points, in cells transduced with adenovirus (http://www.godatabase.org). (Please see Appendix A, Table

A1.) These findings are reasonable considering that BMPs, particularly BMP2, have been

shown to stimulate osteoclast differentiation and activity, possibly by acting on bone marrow stromal cells, in vitro [56]. Thus, the early up-regulation of expression of chemokines such as CXCL2 and genes such as IL-8, which are associated with inflammation may be associated with adenoviral effects, or, more likely, BMP-induced osteoclastic differentiation of BMDMSC. Up-regulation of these genes may also be attributed to an increase in the overall process of regulation of bone production as well as bone turnover, which is accelerated by BMPs.

It is reasonable to believe that our BMDMSC cultures also contained osteoclasts and osteoclast precursors, as it has been shown that bone marrow cultures, in the absence of selection for specific cell types, will contain osteoclasts, osteoclast precursors, and

BMDMSC, and that rh-BMP2 will induce osteoclastic differentiation [56].

Our study appears to be one of the first to report on these early molecular changes in cells transduced with adenovirus. Transduction with adenovirus in vivo is notorious for inciting an inflammatory response [26,63,73] and our results suggest that this is also a cellular event. Adult BMDMSC can be difficult to establish and maintain in culture

[29,84]. They are particularly susceptible to physiologic insults, such as transduction with viral vectors, which can incite inflammatory and apoptotic signaling cascades [26,63].

Therefore, a supportive agent such as BMP-2 may positively affect growth of

BMDMSCs. Transduction with Ad-BMP-2 was able to override this inflammatory/apoptotic effect and sustain cell health and gene expression of hbmp2, as seen by others. However, Ad-BMP-6 was unable to overcome this effect as definitively as Ad-BMP-2, and senescence occurred in the Ad-BMP-6 alginate cultures. Bone

morphogenetic protein-6 was produced in greater concentrations in the monolayer cultures containing Ad-BMP-6-transduced cells, however, protein production was not sustained at the same magnitude as BMP-2 protein by Ad-BMP-2-transduced cells and

BMP-6 was below detection in media from the alginate constructs. In alginate cultures, greater mineralization occurred in the Ad-BMP-6 group, as compared to controls, but we were not able to determine if this was directly BMP-6-mediated. Ad-BMP-6 alginate cultures produced a modest increase in BMP-2, but less than that produced by alginate

BMP-2 cultures. Others have shown that mesenchymal stem cells pelleted and cultured in chondrogenic medium demonstrated greater synthesis of cartilage with BMP-2 than with

BMP-4 or BMP-6 [89]. Based on these data, together with our findings, we conclude that

BMP-2 strongly supported BMDMSC in these in vitro systems.

There are a number of potential explanations for decreased expression of transgene products in alginate cultures, when compared with expression in monolayer cultures. Cells maintained in monolayer culture in our study were not subjected to any physiologic stresses other than removal from 37°C incubation for routine replacement of culture media. In contrast, following transduction in monolayer culture, cells in the 3- dimensional culture group were lifted from monolayers with trypsin, subjected to centrifugation, and resuspended in alginate. This process, which must be carried out at room temperature in order to induce formation of the alginate gels, and involves multiple steps in which the cells are manipulated, resulting in physiologic stress and some degree of unavoidable mechanical trauma to cells. These factors, in combination with other factors, may have effects on transcription, translation, and/or post-translational

modification of transgene products. Additionally these cells are suspended, isolated, and minimally proliferating, which may produce a reduced cellular metabolic rate compared to cells in monolayer. Additionally, the decrease in detectable protein production by the cells within the alginate matrices may be due to protein retention within the alginate matrix. We deduce that transduction of cells used for suspension in alginate cultures was successful because we used the same adenoviral-transgene preparation for cells maintained in both monolayer and alginate cultures.

The expression of a number of genes whose products are relevant to musculoskeletal development, injury, disease, and repair remained unchanged during the course of this study (Appendix A, Table A2). The majority of those gene products were associated with chondrogenic differentiation. In light of our cytochemical and gene expression data, evidence of differentiation of BMDMSC along a chondrogenic lineage did not occur. Our cells demonstrated osteogenic differentiation.

The BMP expression by these equine BMDMSC was able to induce osteogenesis within the alginate constructs despite our use of a nondifferentiating medium, indicating the high relative potency of BMP2 in osteogenesis. Our study is the first to report on the osteogenic differentiation capabilities of BMP-6 on BMDMSC in vitro. In comparison to gene delivery of BMP-2 by similar methods, Ad-BMP-2 induced a more accelerated and robust response than Ad-BMP-6. This may represent greater potency of BMP-2 to induce osteogenic differentiation of BMDMSC, or may reflect the greater BMP production by a healthier cell. In comparison to No Ad, Ad-LacZ, Ad-Luc, Ad-BMP2-transduced

BMDMSC had greater BMP production. This supportive effect may have resulted in the

accelerated osteogenic differentiation of the alginate constructs.

The results presented here demonstrated that high levels of BMP-2 and BMP-6, along with other markers of osteogenic differentiation, can be achieved in BMDMSC in monolayer culture, without the addition of differentiating agents, by adenovirus-mediated gene transduction. This may be valuable in the development of gene therapy approaches to fracture healing. Bone morphogenetic proteins promoted osteogenic differentiation in an even greater capacity in monolayer cultures than in 3-dimensional alginate cultures.

This may be beneficial for injection of BMP-transduced cells into fracture sites. Studies to evaluate the production of BMP-2 and BMP-6 at longer periods post-transduction and use in other models are needed to confirm the findings of this study and to further define the mechanisms of action of BMP-2 and BMP-6 in osteogenic differentiation of pluripotential cells.

The results of this study also demonstrated that alginate may be a suitable 3- dimensional culture system for the delivery of genetically modified BMDMSC. While the response to adenovirus-mediated delivery of growth factors to cells in the alginate system in this study was less robust than the response seen in monolayer cultures, alginate may still be a valuable matrix for the delivery of these cells in in vivo fracture healing studies, where the osteogenic effects of endogenous growth may interact to promote osteogenesis.

The alginate beads used in this study were economical and easy to construct. These cell- alginate constructs could easily be maintained in culture for 22 days, and BMP-2 adequately diffused from the alginate into the media, supporting the hypothesis that similar local BMP release could occur in vivo. An important consideration when

evaluating potential for use in in vivo applications is the ease of handling of this matrix,

particularly the ability to pick up and move the constructs without damaging them. This

characteristic would be of value when planning use of this matrix for delivery of cells to

experimental osteotomies and/or osteochondral defects in animal models.

In summary, transduction of BMDMSC with Ad-BMP-2 or Ad-BMP-6 achieved

osteogenic differentiation, particularly in Ad-BMP-2 cultures. Bone morphogenetic

protein-2 supported osteogenic differentiation of BMDMSC in three-dimensional alginate

cultures, demonstrating the potential utility of this system as a carrier for delivery of

genetically modified cells in vivo. While our results demonstrate that BMP-2 was more robust in vitro, further assessment of BMP-6 or BMP-2/BMP-6 constructs is indicated.

Further studies will be necessary to evaluate the potential of this gene delivery system to augment bone healing in vivo.

aZachos TA, Bertone AL. Adenovirus-mediated bmp2 vs. bmp6 gene delivery to accelerate mineralization in mesenchymal stem cells in monolayer and alginate cultures. Trans Orthop Res Soc, Poster No. 938, 2005

bAdeno-X Rapid Titer Kit, BD Biosciences Clontech, Palo Alto, CA

CHAPTER 3

RODENT MODELS FOR THE STUDY OF ARTICULAR FRACTURE HEALING

3.1 Summary

The goal of this study was to document the healing time course and expression of ex vivo cell-based gene delivery in articular fracture models in the mouse and rat.

Articular medial intercondylar femoral osteotomy was performed in the stifle (knee) joints of hairless immunocompetent mice and medial or lateral similar osteotomy was performed in athymic nude rats. Genetically modified cells expressing a reporter gene product were delivered in a three-dimensional alginate matrix directly into the osteotomy site. Sensitivity of an in vivo imaging system to detect expression of transgene was compared between rodents in this fracture model. Osteotomy healing was assessed using high-detail radiography, helical computed tomography (CT), high-field magnetic resonance imaging, and histology. The mouse model was less satisfactory because the small size of the murine femur made reliable assessment of fracture healing restricted to histopathology and complications occurred in 11/24 mice (45.8%), most frequently transverse supracondylar femoral fracture post-operatively. Gene expression was inconsistently confirmed in mice in vivo for 11 days (p < 0.003). In rats, high-detail

radiography and CT were used to assess osteotomy healing. Magnetic resonance imaging

(4.7T) in rats could produce three-dimensional images that would permit assessment of bone and cartilage, but was time consuming and expensive. In rats, the only surgical complication, transverse femoral fracture, was reduced from 83.3% with the medial osteotomy to 0% with a lateral osteotomy. In vivo imaging confirmed gene expression in the alginate/cell constructs in rats for at least 4 days (p < 0.05). The nude rat model has the advantage of larger femora and the ability to implant xenograft cells. A lateral intercondylar osteotomy of the distal femur in the rat can be used to study the healing of articular fractures.

3.2 Introduction

Articular fractures represent problematic injuries resulting in trauma to bone and surrounding soft tissues in addition to a high risk of long-term complications such as post-traumatic osteoarthritis and subsequent joint incongruity [1-5]. Animal models have been developed to evaluate the mechanisms of fracture healing, sequellae to healing, and to assess various treatments to address these injuries. The majority of published reports on bone repair have studied diaphyseal fractures that do not involve joints [6-10]. Animal models for the study of articular cartilage repair have focused on healing of primary osteochondral defects rather than articular fractures [11-13].

Animal models for the study of articular fractures are not well established. An in vivo model of healing of an articular fracture, particularly one allowing weight-bearing

on the operated limb during fracture healing, would enable the study of the effects of loss of joint integrity on the healing of subchondral bone, articular cartilage, tendons, ligaments, and fibrocartilagenous structures such as the menisci of the knee. The goal of our study was to define bone healing time and pattern in a rodent model of weight- bearing articular fracture, determine if three-dimensional constructs of genetically engineered cells could be detected in vivo, and characterize restrictions on the utility of

outcome measures of healing due to animal size. We serially assess outcome of femoral

intercondylar fracture healing in the mouse and rat using high-detail radiography,

computed tomography, magnetic resonance imaging, in vivo imaging of expression of

reporter transgene products, and histology. Our hypotheses were that osteotomy healing

would occur within a shorter period of time in mice when compared with rats and

osteotomy healing in both species could be reliably assessed using high-detail

radiography, helical computed tomography, in vivo molecular imaging, and histology.

3.3 Methods

Animals

Twenty-four female hairless, immunocompetent albino SKH1 mice (9-11 months of age; Charles River Laboratories, Inc., Wilmington, MA) and 15 female NIH rnu nude rats (10-12 weeks of age; Charles River Laboratories, Inc., Wilmington, MA) were used in this study. This study was approved by the Institutional Laboratory Animal Care and

Use Committee (ILACUC) at The Ohio State University.

Experimental Design

This study used mouse and rat models to investigate articular fracture healing across time with and without a cell/alginate construct. Syngeneic pluripotent mesenchymal stem cells were isolated from bone marrow and genetically engineered to express luciferase. Cells (first, second, and third passages) were suspended in 1.2% alginate for implantation into the osteotomies.

In mice, 11 no-alginate and 13 alginate-treated osteotomies (20 µl n = 7; 10 µl n =

 6) were followed postoperatively using high-detail (Faxitron ) radiography (days 1, 3, 6,

9, 12, 14, and 21) and histology (days 2, 3, 5, 8, 12, and 21). To investigate the detection

of gene expression in cells delivered ex vivo, syngeneic bone marrow-derived

mesenchymal stem cells (BMDMSC) were transduced with a reporter gene (luciferase)

using an adenoviral construct 16 hours prior to suspension in 20 µl (8.9 x 103 - 1.8 x 104 cells, n = 7) or 10 µl (1.8 x 104 cells, n = 6) volumes of 1.2% sodium alginate (ALG)

[14]. Studies performed in our laboratory have confirmed the transduction efficiency of

our adenoviral constructs to be >90% [14].

Post-operative luciferase expression in vivo was serially quantified using an in

 vivo imaging system (IVIS , Xenogen Corporation, Alameda, CA) at days 0, 1, 2, 3, 4, 5,

6, 7, 8, 9, 10, 11, 12, 14, and 21. Luciferase expression of the transplanted BMDMSC was expressed as mean values for flux (photons/sec/cm2/steradian) detected in the

osteotomy stifle joint as compared to contralateral unoperated stifle joints and to

environmental background luminescence. A subset of specimens was harvested for histology on days 2, 3, 5, 8, 12, or 21.

To investigate healing time of the osteotomy in a larger rodent model, 6 rats had a similar medial intercondylar osteotomy as used in the mouse studies. All 6 rats had 5 x

106 BMDMSC from Wistar rats suspended in 50 µl ALG placed in the osteotomy and

 were followed serially by Faxitron radiography at day 28 or day 42, computed

tomography (CT) at days 14, 21, 28, and 42, and histology at day 28 or day 42. To

investigate the detection of luciferase expression in the larger rat model, a subset of 3 rats

in this group had the reporter gene luciferase transduced into 5 x 106 BMDMSCa suspended in the 50 µl ALG construct. These rats were evaluated with in vivo imaging

 (IVIS ) at days 0, 2, and 4.

To investigate whether medial or lateral osteotomy would provide benefits to the

assessment of healing or the model, a lateral intercondylar osteotomy (Figure 1) similar

to the medial osteotomy was created in 9 rats, 3 treated with 50 µl ALG. Healing was

 evaluated using Faxitron radiography, CT, and histology at day 14.

Articular Fracture Model

Mice and rats were pre-medicated with subcutaneous injections of flunixin

meglumine (2.5 mg/kg) and buprenorphine (0.03 mg/kg) for pre-emptive analgesia.

General anesthesia was induced in an anesthesia induction tank using 3% isoflurane in

100% oxygen and maintained by mask. After aseptic preparation, surgery was performed on one hind limb of mice with the assistance of a microscope (Wild/Heerbrugg,

Switzerland) at 6X magnification and on one hind limb of rats with the assistance of 2.5X surgical magnifying loupes. A medial intercondylar osteotomy (24 mice and 6 rats) or a lateral intercondylar osteotomy (Figure 3.11, 9 rats) was performed via medial or lateral parapatellar arthrotomy, respectively, of the stifle joint. After patella luxation, osteotomy of the femoral condyle using a No. 11 scalpel blade was initiated at the articular cartilage of the intercondylar notch and progressed proximo-abaxially to exit the cortex just distal to the proximal attachment of the collateral ligament. Importantly, the collateral ligament was left intact (Figure 1). The osteotomy gap was distracted using the tip of the blade to ensure complete bone fracture of the condyle. The patella was reduced and the skin, subcuticular tissues, and the joint capsule were closed in a single layer using 5-0

 polypropylene sutures (Prolene , Ethicon, Inc., Somerset, NJ) in an interrupted cruciate

pattern in mice. In rats, the joint capsule was closed with 6-0 polypropylene sutures

 (Surgilene , Davis & Geck, Danbury, CT) in a simple interrupted pattern. A continuous

 subcuticular suture line was placed using 6-0 polypropylene suture (Surgilene , Davis &

Geck). All animals were allowed to ambulate freely in their cages immediately postoperatively. Buprenorphine (0.03 mg/kg) was given subcutaneously every 8 hours for 36-

48 hours for post-operative analgesia.

Delivery of Genetically Modified Cells in a Three-Dimensional Matrix

For isolation of syngeneic BMDMSC, bone marrow was harvested from the femora and tibiae of mice. Marrow was flushed into 60 mm x 15 mm Petri dishes containing Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 20% fetal bovine serum, 50 units of sodium penicillin/ml, 100 units of streptomycin/ml, and 29.2 units of L-glutamine/ml (Gibco, Grand Island, NY). The same procedures were performed to isolate and expand BMDMSC from Wistar rats. For a subset of rats, allograft BMDMSC were obtained from male Lewis ratsa.

Cells were expanded in monolayer culture and a subset of cells was transduced

with a first-generation adenoviral-luciferase transgene construct as previously described

[14, 15]. Cells were then suspended in 10, 20, or 50 µl alginate gels using a previously described protocol [14, 16]. At osteotomy surgery, one cell-alginate construct was placed into each distracted osteotomy site using a No. 0 bone curette and a 22g needle to manipulate the construct.

In Vivo Imaging of Reporter Gene Product

Imaging of both operated an unoperated distal femora was performed to quantify duration and intensity of gene (luciferase) expression using an in vivo imaging system

 (IVIS , Xenogen) at days 0 (immediately post-operatively), 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,

12, 14, and 21 post-operatively. Quantification of transgene expression was accomplished

by comparing values for flux acquired by a charged cooled coupled device (CCCD) camera, in the operated stifle to the unoperated contralateral stifle, and between animals receiving AdLuc-transduced cells in alginate and animals receiving ALG constructs

 containing untransduced cells or no cells. Imaging using the IVIS system was performed

by intraperitoneal injection of 150 mg/kg of luciferin, the substrate for the reaction

yielding luciferase, 5-15 minutes prior to imaging under general anesthesia, in both mice

and rats.

Radiography

Fracture healing was evaluated using craniocaudal and mediolateral radiographs

 of the mouse femora using a Faxtiron Cabinet X-Ray System (Faxitron X-Ray

Corporation, Wheeling, IL). Radiographs were taken at days 1, 3, 6, 9, 12, 14, and 21

post-operatively. An exposure time of two minutes and 30 seconds at 35 kilovolt peak

(kVp) was used for mice. An exposure time of 3 minutes at 35 kVp was used for rats.

Similar radiographs were taken immediately following euthanasia at postoperative days

14, 28, and 42. An exposure time of three minutes and 30 seconds at 35 kVp was used in

rats.

Computed Tomography

Computed tomography (CT) was performed following euthanasia on day 12 in 5

mice using a Picker PQS helical CT scanner, (Phillips Medical Systems, Bothell, WA) in

the Veterinary Teaching Hospital at The Ohio State University. Imaging parameters were

130 kVp and 125 milliamperes per second (mAs) with a bone window field of view

(FOV) of 180 mm. Tomographic slice thickness/width was 1 mm/1 mm. In rats, CT was performed at days 14, 21, and 28, and 42 under general anesthesia using the same CT scanner. Imaging parameters were 130 kVp and 100 mAs with a field of view of 13.0 x

13.0 cm. Slice thickness/width was 1 mm/1 mm.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) of the mouse stifle joint was performed on day 21 following euthanasia using a 4.7 Tesla/40 cm magnet controlled by a Bruker

Avance console with a 120 mm inner diameter gradient coil (max. 400 mT/m; 170 microseconds rise time; diameter of spherical volume: 8.0 cm; shielded, water cooled) with a 72 mm inner diameter proton volume radiofrequency coil.

To determine the feasibility of MRI of rat stifles, proton density (PD)-weighted images of the rat stifle joint were obtained in the coronal, sagittal, and axial imaging planes. Proton Density (PD) weighted images were acquired using a spin echo (SE) sequence with a repetition time (TR) = 2500 ms and echo time (TE) = 6.9 ms. Field of view (FOV) and matrix size for the 10 sagittal images were as follows: FOV = 4.0 x 4.0 cm2, matrix = 256 x 256. Acquisition time for this sequence is 10 minutes, 40 seconds.

For coronal and axial data, this time was reduced by halving both FOV and matrix sizes

in the phase encode directions.

Histologic Analysis

Osteotomized limbs were harvested at time 0 (i.e. immediately following osteotomy) in cadavers, and at days 2, 3, 5, 8, 12, and 21 following euthanasia in mice and at days 14 and 28 in rats. Euthanasia was performed by CO2 inhalation. Limbs were harvested and fixed in 10% neutral buffered formalin (NBF) for 72 hours at room temperature. The limbs were decalcified in 10% ethylenediaminetetraacetic acid (EDTA) for 72 hours (mouse limbs) or 21 days (rat limbs). The limbs were placed in 10% NBF for transport to the histotechnology facility, where they were dehydrated in a series of alcohols, embedded in paraffin, cut into 6-µm sections, stained with hematoxylin and eosin or safranin O and fast green, and examined using light microscopy.

Statistical Analysis

Objective data obtained were compared between control and osteotomy femurs

 for IVIS and between healing bone and adjacent bone for CT using the Kruskal-Wallis

 test and Dunn’s post test for multiple comparisons (GraphPad PRISM 4, GraphPad

Software, Inc., San Diego, CA). A p-value of < 0.05 was considered significant for all

analyses.

3.4 Results

Clinical Outcome of Surgical Procedure

Operative time for the surgical procedure, including creation of the articular osteotomy, was less than 15 minutes. Complications are summarized in Table 1. In mice, complications occurred in 11/24 mice overall (45.8%) and included transverse supracondylar fractures in 4/24 mice (16.7%) occurring within 48 hours post-operatively.

Lateral patellar luxation occurred in 4/24 mice (16.7%), and was diagnosed based on radiographic findings between post-operative days 3 and 9. Physeal fractures occurred intra-operatively in 3/24 mice (12.5%). In rats, five of the 6 (83.3%) undergoing medial intercondylar distal femoral osteotomies had similar transverse supracondylar fractures within 48 hours postoperatively. None of the rats with lateral intercondylar femoral osteotomies had physeal fracture. No complications occurred with the lateral intercondylar osteotomy model.

In Vivo Imaging of Reporter Gene Product

Luciferase expression was greater until day 11 in mice (p < 0.003) and day 4 in

rats (p < 0.05) in joints receiving AdLuc-transduced BMDMSC in alginate when

compared with joints receiving untransduced cells in alginate. Values for expression in

mice are shown in Figure 3.2.

Radiography

Sixty percent of untreated osteotomies were healed, as evidenced by bony bridging, at, but not before, day 9. In the alginate-treated osteotomies, healing was identified in 1 mouse at day 12 and 33% of mice by day 21.

In mice, images obtained at day 0 were useful in providing radiographic confirmation that the osteotomy procedure was successful in creating a complete, displaced fracture in the medial cortex of the distal femoral metaphysis (Figure 3.3). No appreciable narrowing of the osteotomy gap was consistently noted until day 14. At day

21, a thin layer of periosteal new bone was consistently seen spanning the medial cortex of the femur, from the level of the mid-diaphysis to the proximal aspect of the medial femoral condyle, but bony bridging of the fracture was not completed by day 21 (Figure

3.3).

In the rat model, medial osteotomies all had transverse supracondylar fractures which were completely healed and remodeled by day 42. Supracondylar fracture was eliminated with use of the lateral osteotomy. Fifty percent of lateral supracondylar osteotomies were healed radiographically by 14 days (Figure3. 4).

Computed Tomography

Axial computed tomography (CT) images demonstrated sites of periosteal new bone formation in some mice by day 12. Evidence of osteotomy healing was not consistently demonstrated, and was difficult to evaluate with the resolution of standard

CT (1 mm slices) due to the small size of the mice femora. In rats, osteotomy sites could be discerned as incomplete bony bridging at days 21 and 28 (Figure 3.5), however, significant differences in bone mineral density could not be resolved between the osteotomy site and adjacent bone or between days 14, 21 and 28 with standard CT (p >

0.05).

Magnetic Resonance Imaging

Magnetic resonance images were obtained in the axial (data not shown), coronal,

and sagittal planes in both the mouse and rat models. The fracture gap with associated

callus was most easily visualized in the coronal and sagittal images (Figure 3.6). An

interesting finding on the sagittal MRI of the mouse femur was that the periosteal new

bone, which was not reliably detected with high-detail radiography, was seen at day 21

with MRI. Duration of time necessary to obtain MRI was approximately 7.5 hours in the

mouse. In the rat stifle, the trabecular pattern of metaphyseal and epiphyseal bone, as

well as the physes, collateral ligaments, menisci, infrapatellar fat pad, patellar ligament,

and articular cartilage were clearly visualized and well matched with histology (Figure

3.7). Duration of time necessary to obtain MR images was approximately 16 hours in the

rat.

Histologic Analysis

In sections of mouse femora stained with hematoxylin and eosin, osteotomy sites demonstrated evidence of healing by endochondral ossification at day 12 (Figure 8).

Organizing fibrous tissue was also noted in all healing osteotomy sites. Consistent with

 findings obtained using Faxitron radiography and CT, evidence of bony bridging of the

osteotomy gap was inconsistently seen at this time point. In ALG-treated osteotomies a

round, well-delineated mass of cells resembling chondrocytes in lacunae were seen in the

healing osteotomy site at days 12 and 21 (Figures 3.8B and C).

In the rat model, histologic specimens from the medial osteotomy group revealed

partial healing of osteotomy sites by endochondral ossification, with varying amounts of

fibrous tissue and cartilage, by day 14, and partial bony bridging at day 28.

Rat femora in the modified lateral osteotomy group at day 14, osteotomies treated

with alginate had evidence of fibrous tissue organization, early mineralization and in

some cases, organized fibrous tissue and cartilage, consistent with endochondral

ossification (Figures 3.9A and B).

3.5 Discussion

The rodent osteotomy model described here permitted the creation of an articular

fracture, which served as a tool to assess articular fracture healing. The lateral condylar

articular fracture permitted repeatable articular osteotomy without musculoskeletal

complications. Supracondylar fracture and other musculoskeletal complications limited

the use of this model in mice and on the medial condyle of rats. Other limitations of the model in mice include inability to accurately assess osteotomy healing by high-detail radiography or conventional CT and the impracticality of MRI currently to image joints this small. Bone quality varies with mouse strain and may influence the likelihood of fracture complications [17-19]. In vivo imaging of reporter gene expression in the mouse model was possible, but inconsistent and likely due to the smaller numbers of cells.

Luciferase expression was weak and hard to resolve from background non-specific photon emission from adjacent tissue. The result was inconsistent identification of luciferase expression with this mouse model.

In vivo imaging of reporter gene expression in the rat model was successful in providing a reliable method to easily discern differences between luciferase expression in femora containing cells expressing this transgene product and unoperated controls. The rat model permitted the use of a greater number of cells in each construct, which resulted in more consistent identification of luciferase expression in vivo and immediately upon

surgical implantation of the cell-alginate constructs. This technique served to confirm

retention of the construct at the surgery site and initial survival of cells capable of

transgene expression. Detecting gene expression in bone is difficult by any technique due

to the required processing of mineralized specimens. Our model challenges this

bioluminescent technology [20] which was designed for the quantification of luminescent

reporter gene expression in animal models with tumor metastases. Metastatic cancer

lesions are masses of cells with a high metabolic rate and frequently located in

homogeneous tissues making detection of expression less susceptible to scatter and attenuation [20]. In our model, we used adult BMDMSC of modest metabolic rate suspended in isolation in a three-dimensional alginate matrix. The heterogeneity of mineralized bone in this anatomic location posed a considerable challenge to confirming transgene expression in vivo.

 The utility of Faxitron radiography was limited as a tool for evaluation of

healing in the mouse model and could not consistently determine healing time of the

fracture. Radiographs could be used to document osteotomy configuration and position

only. These limitations were attributed to the small size of the specimens. The

intercondylar width of the mouse femora (measured from the lateral epicondyle to the

medial epicondyle) was approximately 3 mm, versus approximately 5.5 mm for the rat

femora. The use of CT at the resolution obtained with the helical scanning system (350

µm voxel resolution for mouse femora, 250 µm voxel resolution for rat femora) was also

not adequate for the assessment of healing of articular osteotomies in mice. While

resolution was better in the larger rat femora, this modality lacked the sensitivity required

for discerning differences among fractured and non-fractured areas in the rat model.

Imaging with higher spatial resolution CT, such as micro-CT imaging, would be

recommended for either rodent model [21-23]. Magnetic resonance imaging provided

high-detail imaging of the bone and the cartilage and other soft tissues of the joint. This

advantage of MRI over CT is significant. The design of alternative coils [24] that can

scan small specimens in a shorter time may be able to provide substantial improvement

over CT, including micro-CT, in quantifying three-dimensional bone and cartilage healing of articular fractures in rodents. High-field MRI had the greatest potential to correlate with bone and cartilage healing seen with routine histopathologic sections.

High-field MRI, using a system designed for use in small laboratory animals, was ideal for the evaluation of not only the intra-articular and peri-articular soft tissues of the stifle joint, but also of the subchondral bone of the epiphyses and the distal metaphyseal trabecular bone. High-field MRI is a valuable modality for the evaluation of joints of mice and rats, and is worthy of further study for use in in vivo rodent models.

We describe a successful rodent model of articular fracture that can be used to

study cell and gene delivery for bone healing. In the rat, a lateral condylar osteotomy

avoided complications. The rat model provided greater consistency and resolution of

healing with CT and MRI than the mouse model. Alginate served as an acceptable

delivery vehicle for BMDMSC into the osteotomy site. Histology confirmed early

osteotomy healing with evidence of endochondral ossification between days 12 and 21

 for mice and by day 42 for rats. Faxitron radiography did not provide sufficient detail to

reliably confirm osteotomy healing. The rat model described here is well-suited to study

articular fracture healing and the evaluation of gene delivery as an investigational

modality to accelerate fracture healing.

aBone marrow-derived mesenchymal stem cells (BMDMSC) from male Lewis rats (generously provided by Dr. Darwin Prockop, Roxanne Reger, MS, and Margaret Wolfe, MS of the Tulane Center for Gene Therapy through a grant from NCRR of the NIH, Grant # P40RR017447)

cranial

Figure 3.1: Illustration of the rat stifle joint depicting the lateral intercondylar osteotomy model. (Image by Tim Vojt)

Luciferase Expression: Mice 8000 AdLuc-BMDMSC Control 7000

6000

5000

Flux 4000

3000

2000

1000

0 Day 2 Day 4 Day 6 Day 8 Day 10 Day 11

Figure 3.2: Luciferase expression in distal femora of mice. Control mice were those receiving untransduced BMDMSC in alginate. Flux values (photons/sec/cm2/sr) were significantly greater for AdLuc-BMDMSC mice than for controls (p < 0.003). Bars represent standard errors of the means..

Day 0 Day 21

Figure 3.3: Faxitron high-detail radiographs taken of mouse stifle joints at days 0 and 21. The arrow indicates the proximal aspect of a layer of periosteal new bone first documented at this time point, demonstrating progressive healing during these time points, but incomplete bony bridging by day 21.

Figure 3.4: Faxitron high-detail radiographs taken of two stifle joints from athymic nude rats (NIH rnu) at day 14. These lateral intercondylar osteotomies had not healed (arrows) by this time point.

Figure 3.5: Axial computed tomographic (CT) images of healing intercondylar osteotomies in NIH rnu (athymic nude) rats at day 21. The small size of the femora made serial evaluation of osteotomy healing difficult using a 1 mm slice thickness. The unhealed osteotomy site can be appreciated in the left femur (arrow).

Figure 3.6: Proton density-weighted magnetic resonance images (MRI) of healing intercondylar femoral osteotomy at day 21 in a mouse. The fracture callus at the proximal aspect of the medial femoral condyle is indicated by the green arrow in the coronal image (A). In the sagittal image (B), the fracture callus at the proximal aspect of the medial femoral condyle is indicated by the blue arrow. The bracket indicates the periosteal new bone seen at later time points in radiographs (Figure 3). (MRI parameters: 3D SE, TR/TE = 2500/5.7 ms, in-plane resolution = 156 x 156 µm2; 2D SE, TR/TE = 2500/6.9 ms)

Figure 3.7: Proton density (PD)-weighted MRI of unoperated control NIH rnu rat stifle joint obtained with 4.7T magnet showing trabecular bone, physes, articular cartilage, and articular and periarticular soft tissue structures. A Coronal imaging plane. B Sagittal imaging plane. (MRI parameters: 3D SE, TR/TE = 2500/5.7 ms, in-plane resolution = 156 x 156 µm2)

A B C

Figure 3.8: Light microscopic appearances of medial intercondylar femoral osteotomies in SKH1 (immunocompetent) mice. A Osteotomy site (arrow) immediately following procedure (time 0). (H&E X100) B Healing osteotomy site, day 12, untreated osteotomy. An irregularly shaped cartilage mass (arrowhead) and adjacent organizing fibrous tissue (arrow) are present in the healing osteotomy site. (H&E X200) C. Healing osteotomy site, day 12. Syngeneic BMDMSC were delivered to the osteotomy site in this animal in alginate. The dashed line indicates the margins of a round, compact cartilage mass, separated from any organizing fibrous tissue, at the osteotomy site. (H&E X100) Inset: High-power view of chondroid tissue mass. (H&E X400)

B A

Figure 3.9: Light microscopic appearances of lateral intercondylar femoral osteotomies in NIH rnu (athymic nude) rats on day 14. A Untreated osteotomy. Fibrous tissue (arrows) is present at the osteotomy site. (H&E X100) B Osteotomy site treated with empty 50 l alginate bead demonstrating the presence of a round mass of tissue (dashed line) in the osteotomy site. (Safranin O/fast green X100)

Complication Number of Mice Affected Physeal fracture 3 Transverse 4 supracondylar fracture Lateral patellar 4 luxation

Table 3.1: Complications associated with creation of medial articular osteotomies in mice.

CHAPTER 4

CHONDRO-OSSEOUS DIFFERENTIATION OF BONE MARROW-DERIVED MESENCHYMAL STEM CELLS IN ALGINATE CULTURES INDUCED BY BONE MORPHOGENETIC PROTEIN-2 AND-6 GENE DELIVERY

4.1 Summary

Bone marrow-derived mesenchymal stem cells (BMDMSC) are pluripotential cells which can undergo chondrogenic and osteogenic differentiation. Delivery of genes encoding chondrogenic and osteogenic growth factors to BMDMSC has been proposed as a method for development of cell-based therapies for bone and cartilage regeneration.

We evaluated the potential of a three-dimensional alginate system for the controlled chondro-osseous differentiation of BMDMSC from Lewis rats in vitro. We used a first- generation recombinant human adenoviral vector to deliver the cDNAs encoding human bone morphogenetic proteins-2 and -6 (AdBMP2, AdBMP6) to cells prior to suspension in alginate. Untransduced BMDMSC in alginate served as controls. Results of quantitation of gene expression and immunohistochemistry were consistent with chondro-osseous differentiation of cells in all groups. However, the most robust chondrogenic response was seen in the cells transduced with AdBMP2 and

mineralization of cells was delayed. The AdBMP6 constructs were smaller, of greater cell density, and mineralized early, representing osteogenic differentiation. Chondrogenic differentiation of BMDMSC can be achieved in AdBMP2 cell-alginate constructs, and chondro-osseous differentiation of these pluripotent cells can be achieved in untransduced constructs and AdBMP6 constructs, without the use of differentiating agents in specialized culture media.

4.2 Introduction

Osteochondral defects present a significant clinical challenge. Articular cartilage is an avascular, aneural tissue and heals poorly. Numerous tissue engineering approaches have been developed, in which a three-dimensional matrix or scaffold is used as both a supportive environment and a delivery system for cells to promote cartilage healing.

Gene delivery to chondrocytes and bone marrow-derived mesenchymal stem cells

(BMDMSC) has served as a mechanism [1-10] to allow for the augmentation of tissue- engineered bone and cartilage replacement systems.

One of the delivery vehicles which has been well-studied to date is alginate, a biodegradable linear copolymer derived from the kelp plant (Macrocystis pyrifera).

Alginate is economical to purchase and store. It also has the important advantage of ease

of handling, which becomes relevant when looking toward translational applications.

While the differentiation of BMDMSC in alginate has been studied in vitro [11-16] and in vivo [17-20], studies published to date have used chondrogenic and/or osteogenic

differentiating media containing growth factors known to induce the desired differentiation status of BMDMSC. These studies have consistently demonstrated that

BMP2 [11,12,15,16,18,19,20] and other members of the BMP subfamily of growth factors in the transforming growth factor-β (TGF-β) family of proteins

[12,13,14,15,16,19] induce chondrogenic differentiation of BMDMSC via the mechanism

of endochondral ossification in alginate. However, in these studies, cells were pre-treated

with and maintained in chondrogenic differentiating media to achieve these results. In

some cases, instead of using a population of cells isolated from bone marrow, cell lines

known to be capable of differentiation into chondrocytes, were used [15,16,18].

Delivery of cDNAs encoding sonic hedgehog [21] bone morphogenetic protein

(BMP)-7 [22], BMP-2 [19,23-33], and BMP-6 [34-35] have been described, either via ex

vivo gene delivery to periosteal cells and BMDMSC in alginate and other three-

dimensional scaffolds, or via direct gene delivery in in vivo models. Bone marrow-

derived mesenchymal stem cells are attractive targets for both in vivo and ex vivo gene

delivery systems, as they can differentiate, in response to various molecular signals, into

chondrocytes and osteoblasts, among other cell types [3,36-38]. In addition, these cells

have the advantage of being readily available for autologous harvest via minimally

invasive approaches, and can be expanded in vitro. Others have demonstrated the ability

to induce differentiation of BMDMSC into constructs containing both osteoblasts and

chondrocytes. However, to date this directed chondro-osseous differentiation of

BMDMSC has only been accomplished when the cells were maintained in chondrogenic and osteogenic differentiating media and/or treated with chondrogenic and osteogenic growth factors in the form of recombinant proteins [1,4-5].

The goal of our study was to evaluate the in vitro differentiation of BMDMSC suspended in a biologic three-dimensional matrix following delivery of the cDNA encoding an ostoeochondrogenic growth factor. Our hypothesis was that AdBMP2 constructs would undergo chondrogenic differentiation, AdBMP6 constructs would undergo osteogenic differentiation, and that untransduced cells constructs would not undergo chondrogenic differentiation.

4.3 Results (Note: Methods section begins on page 87.)

4.3.a RT-PCR Analysis of Gene Expression

The level of hbmp2 gene expression at day 17 in the AdBMP2 constructs was 11-

fold greater when compared with 18SrRNA expression and normalized to untransduced

constructs at the same time point. Expression of hbmp6 was not detected in these

AdBMP2 constructs. The level of col2a1 gene expression at day 17 in the AdBMP2

constructs was 287-fold greater when compared with 18SrRNA expression and

normalized to untransduced constructs at the same time point. The level of aggrecan

(aggr) gene expression at day 17 in the AdBMP2 constructs was 5190-fold greater when

compared with 18SrRNA expression (Table 4.2). The level of col1a2 gene expression at

day 17 in the AdBMP2 constructs was 740-fold greater when compared with 18SrRNA

expression (Table 4.2). The ratio of induction fold increases in gene expression of col1a2

to col2a1 at day 17 in AdBMP2 constructs, relative to 18SrRNA expression and normalized to untransduced control constructs at the same time point, was 2.6.

4.3.b Histologic Evaluation

Size of Alginate Constructs

Cell-alginate construct size demonstrated that at day 17, constructs containing untransduced constructs were significantly larger than alginate constructs without cells

(EMPTY ALG; p = 0.05). At day 22, the size of the AdBMP2 constructs was not significantly different from that of the AdBMP6 constructs. At day 22, untransduced constructs were significantly larger than AdBMP6 constructs (p < 0.05), but were not significantly larger than AdBMP2 constructs (p > 0.05; Table 4.2).

Cell Scoring

Median cell density scores for the day 22 AdBMP6 constructs were significantly greater than those for the day 22 untransduced constructs and for the day 22 AdBMP2 constructs (p < 0.02; Table 4.2). The BMP6 constructs were smaller, had a greater cell density, and underwent osteogenic differentiation early and more readily than untransduced and AdBMP2 constructs (Figure 4.2; Table 4.2).

4.3.c Immunohistochemistry for Type II Collagen

The AdBMP6 constructs stained with significantly less intensity and to a lesser extent than the AdBMP2 and untransduced constructs for type II collagen (Figure 4.3;

Table 4.2). Immunohistochemistry specimens were evaluated using light microscopy

(Figure 2). The day 22 untransduced constructs had a median score of 3.25 for intensity of type II collagen immunostaining and a median score of 2.5 for extent of staining. The day 22 AdBMP2 constructs had a median score of 2.5 for intensity of type II collagen immunostaining and a median score of 3 for extent of staining. The day 22 AdBMP6 constructs had a median score of 1.5 for intensity of type II collagen immunostaining and a median score of 1 for extent of staining. The AdBMP2 constructs were larger and demonstrated chondrogenic morphology without evidence of mineralization by day 22, suggesting sustained chondrogenic phenotype.

4.4 Discussion

In the current study, we demonstrated that BMDMSC from adult Lewis rats in primary culture undergo chondrogenic differentiation as a default pathway when maintained in alginate as a standard, non-differentiating three-dimensional culture medium in vitro. Mineralization was noted by day 6 in untransduced constructs and was not noted in AdBMP2 constructs at any time point evaluated. This co-differentiation process appears to parallel the process of endochondral ossification, as it would occur in

vivo in BMDMSC transduced ex vivo with the gene encoding human BMP-2 [6,31-

33,38]. We anticipated that transduction of BMDMSC with the gene encoding human

BMP-2 prior to suspension in alginate would result in significantly greater chondrogenic

differentiation of cells, as has been shown by others using directed chondrogenic media

[7-10, 36-48]. However, in our study, chondro-osseous differentiation of cells in a

biologic three-dimensional matrix was achieved without the use of chondrogenic and

osteogenic differentiating media. Indeed, robust up-regulation in gene expression of

hbmp2, aggrecan, and col1a2, relative to 18SrRNA expression by day 17 in the AdBMP2

constructs is consistent with chondrogenic induction. However, the greater relative

increase in col1a2 gene expression at the same time point suggested that a simultaneous

osteogenic differentiation process was occurring. Evaluation of gene expression at serial

time points prior to day 17 may be valuable in the elucidation of the order of upstream

mechanisms of chondro-osseous induction, specifically the identification of the relative

time frame of chondrogenic induction and osteogenic induction, in this in vitro system.

Specifically identification of a time frame during which chondrogenic induction becomes osteogenic induction may provide investigators with valuable information concerning regulation, and ultimately, manipulation of this process in vitro, with the goal of implementing this knowledge in in vivo applications. The in vitro evaluation time points of days 17 and 22 were chosen based on the results of previous in vitro studies in which

osteogenic differentiation of BMDMSC in alginate was demonstrated in response to

BMP gene delivery without the use of pre-differentiating conditions [44].

A number of morphological differences were found on evaluation of histologic sections between cell-alginate constructs in the AdBMP2 and AdBMP6 constructs and those in the untransduced constructs at day 22. Our numeric scoring systems did not demonstrate statistically significant differences between groups, specifically with regard to cell density, degree of staining with toluidine blue or von Kossa, and cell distribution.

However, one of the most striking differences consistently noted between the groups was that while some degree of mineralization was noted in both the untransduced constructs and the AdBMP6 constructs at by day 22, no mineralization was seen in the AdBMP2 constructs, suggesting a delay in mineralization, compared to the other groups. Up- regulation of type I collagen gene expression suggests cell molecular signaling was compatible with osteogenic differentiation. In this three-dimensional alginate culture system, BMP2 gene delivery supported the acceleration and persistence of chondrogenic differentiation as compared to AdBMP6- or untransduced constructs.

Morphology and histochemical staining supported the chondrogenic differentiation in the AdBMP2 constructs, cells were larger and rounded, surrounded by lacunae, and separated by an extracellular matrix which stained with toluidine blue, consistent with chondrogenic differentiation. Cells in the untransduced constructs were smaller and disorganized within the alginate constructs with minimal toluidine blue staining. Cells in the AdBMP6 constructs were also small and appeared to form swirling palisades with minimal extracellular matrix, and greater mineralization than cells in the untransduced and AdBMP2 constructs. The untransduced and AdBMP2 constructs

overall had more diffuse and more intense toluidine blue staining than AdBMP6 constructs. Unlike the untransduced constructs and the AdBMP6 constructs, the AdBMP2 constructs demonstrated no positive von Kossa staining. Extracellular mineralization, which is necessary but not sufficient for the documentation of osteogenic differentiation, was noted in the untransduced and AdBMP6 constructs, but not in the AdBMP2 constructs.

Similarly, demonstration of type II collagen production is consistent with chondrogenic differentiation of cells, most particularly in AdBMP2 constructs immuno- type II staining was minimal in the AdBMP6 constructs, and extensive in both the untransduced and AdBMP2 constructs. These results suggested that when rat BMDMSC are suspended in alginate the default differentiation pathway was chondrogenic differentiation and endochondral ossification.

Spontaneous chondrogenic differentiation of BMDMSC has not been seen in other three-dimensional constructs, and may be unique to alginate. Others have demonstrated chondrogenic differentiation of BMDMSC in alginate [11-16], fibrin [40] and agarose [41,42]. However, in all of these cases, the cells in the three-dimensional constructs were treated with differentiating agents known to specifically induce differentiation along the chondrogenic lineage. Our study is unique in that we achieved chondrogenic differentiation of cells in alginate with neither pre-treatment of cells with nor maintenance of the constructs in differentiating media. This finding, and its serial confirmation to day 22 in culture, has not been previously reported. This suggests that in

alginate, the default pathway for these particular BMDMSC in standard media in this in vitro system, is chondrogenic differentiation. In addition, AdBMP2 gene augmentation of

BMDMSC in alginate resulted in persistence of the chondrogenic genotype and phenotype. The process of endochondral ossification may be controlled in vitro by the delivery of genes encoding growth factors such as BMPs. These findings may be valuable in the development of biologic cell-based delivery systems for evaluating reconstruction of osteochondral defects in in vivo models of bone and cartilage healing.

In summary, our findings suggest that in the three-dimensional alginate system evaluated here, BMDMSC from immunocompetent rats will undergo chondrogenesis.

This response is most robust and persistent with AdBMP2 gene augmentation.

Adenoviral-BMP6 gene augmentation promoted cell proliferation and mineralization resulting in osteogenic differentiation. Our study uniquely demonstrated chondro-osseous differentiation of BMDMSC when alginate is used without the use of chondrogenic and osteogenic differentiating media.

4.5 Methods

Generation of adenoviral vector constructs

Recombinant adenoviral vectors containing either a 1547 base-pair open reading frame segment of human BMP-2 [43] or a 1539 base-pair open reading frame segment of

human BMP-6 [34], under the control of the cytomegalovirus promoter, were generated according to the methods described by Bertone et al [44]. Expression of transgenes was verified in cell culture [34,44].

Three-dimensional culture system

Bone marrow-derived mesenchymal stem cells (BMDMSC) from male Lewis rats were obtained and were expanded in monolayer culture under standard culture conditions

(exactly as specified in, “Culture of Rat Marrow Stromal Cells (rMSC),” provided by

Tulane Center for Gene Therapy with shipment of cellsa). When monolayer cultures had

reached approximately 70% confluence, transduction of cells with an adenoviral-BMP2

transgene construct (AdBMP2) or an adenoviral-BMP6 transgene construct (AdBMP6)

was performed at an infectious multiplicity of infection of 8.5:1b at 37°C for a

transduction time of two hours, washed and allowed to incubate overnight to achieve

expression of transgene. Untransduced cells (passages 7-9) suspended in alginate

(untransduced constructs) and alginate constructs containing no cells (EMPTY ALG)

served as controls. The transduction efficiency of these adenoviral constructs was

previously determined to be >90% [44].

Monolayers were rinsed with 1X calcium- and magnesium-free phosphate-

buffered saline (PBS, Gibco, Grand Island, NY) and treated with 0.25% trypsin-EDTA

(Gibco) for five minutes at 37°C. Alpha-Minimum Essential Medium (α-MEM, Gibco)

supplemented with 20% fetal bovine serum, sodium penicillin at a concentration of 100

units/ml, streptomycin at a concentration of 100 units/ml, and L-glutamine at a concentration of 58.4 mg/ml was added to the monolayers to terminate the trypsin reaction. The cells were subjected to centrifugation at 450 x g for 8 minutes and

resuspended in a 1.2% solution of sterile sodium alginate (PRONOVA SLG100,

NovaMatrix, FMC Biopolymer, Oslo, Norway) at a concentration of 5 x 106 cells per 50

µl volume. Aliquots of the cell-alginate suspension in 50 µl volume were placed into a

102 mM solution of CaCl2. Constructs were solidified in the CaCl2 solution for 10

minutes, rinsed three times with sterile 0.9% NaCl, and each 50 µl construct was placed

in one well of a 48-well tissue culture plate (Discovery Labware, BD Biosciences,

Bedford, MA) containing α-MEM supplemented as described above, and incubated at

37°C with 5% CO2.

RT-PCR Analysis of Gene Expression

Gene expression of human bone morphogenetic protein 2 (hbmp2), human bone

morphogenetic protein 6 (hbmp6), the α1 chain of type 2 collagen (col2a1), aggrecan

(aggr), and the α2 chain of type 1 collagen (col1a2) were quantified using real-time reverse-transcription polymerase chain reaction (RT-PCR) using the ABI PRISM 7000

Sequence Detection System (Applied Biosystems, Foster City, CA). Cell-alginate constructs harvested at day 17 post-gel construction were disrupted using a tissue homogenizer. Total RNA from cells was isolated by phenol-chloroform extraction

 (TRIZOL Reagent, Invitrogen Corp., Carlsbad, CA). Mean fold changes in hbmp2,

hbmp6, col2a1, aggr, and col1a2 gene expression were calculated, relative to expression

-∆∆C of the endogenous control gene encoding eukaryotic 18SrRNA, using the 2 T method

[45]. For comparison, the CT values of the untransduced (NoAd) cell-ALG constructs

were used as normalized controls when available [46]. The primer and probe sets for

hbmp2, hbmp6, and col2a1 were designed by Applied Biosystems as part of the

 commercially available Assays on Demand TaqMan Gene Expression Assays service

(Table 4.1). The primers and probes for aggr and col1a2 were designed using Primer

 Express software v2.0 (Applied Biosystems).

Histologic Evaluation

At days 0, 1, 4, 6, 17, 22, untransduced, AdBMP2, AdBMP6, constructs were

harvested and fixed in 70% ethanol overnight and transferred to 10% neutral buffered

formalin. They were processed for routine histologic evaluation by dehydration in a

series of graded alcohols, embedded in paraffin, cut into 5 µm-thick sections and

mounted on glass slides, and stained with hematoxylin and eosin, toluidine blue, and the

von Kossa method.c The size of the cell-alginate constructs was determined from fixed, mounted specimens by measuring the width of the constructs at both their widest and narrowest points. These values (mm) were multiplied to arrive at a value for area (sq mm) for each construct.

A numeric scoring system was used to evaluate cell density, degree of staining with toluidine blue, degree of von Kossa staining, and cell distribution. Cell density for each specimen was given a score of 0-4 with a score of 0 representing the least cellular specimen and a score of 4 representing the most cellular. Toluidine blue staining for each specimen was given a score of 0-4 with a score of 0 representing no staining and a score of 4 representing the most intense staining. Cell distribution for each specimen was given a score of 0-4 with a score of 0 indicating that cells were confined to periphery and a score of 4 indicating that cells were uniformly distributed throughout construct. Degree of von Kossa staining for each specimen was given a score of 0-4 with a score of 0 representing no staining and a score of 4 representing the most intense staining. In all cases, the average of scores from two observers was calculated and expressed as means.

Immunohistochemical Staining Procedure for Collagen II

Immunohistochemistry for type II collagen was performed on a subset of constructs at day 22. Paraffin embedded tissue was cut at 4 µm and placed on positively charged slides. Slides with specimens were then placed in a 60°C oven for 1 hour,

cooled, and deparaffinized and re-hydrated through xylenes and graded ethanol solutions

to water. All slides were quenched for 5 minutes in a 3% hydrogen peroxide solution in

water to block for endogenous peroxidase.

Antigen retrieval was performed by a heat method in which the specimens were placed in Dako’s Target Retrieval Solution (pH 6.1) for 25 minutes at 94°C in a vegetable steamer and cooled for 15 minutes in solution. Slides were then placed on a

Dako Autostainer, (Dako Cytomation California, Inc., Carpinteria, CA) immunostaining system, for use with immunohistochemistry. Slides were first protein blocked with

Dako’s serum-free block for 15 minutes, then collagen II antibody was applied at a 1:2 dilution and incubated for 60 minutes at room temperature. Next slides were blocked for endogenous biotin and the secondary antibody applied was goat anti-mouse (Vector, BA-

9200), at 1:200 for 30 minutes. The Vectastain Elite (Vector, PK-6100) detection system was used for 30 minutes. Slides were developed with DAB chromogen for 5 minutes.

Slides were then counterstained in Richard Allen hematoxylin, dehydrated through graded ethanol solutions and cover slipped.d

Degree of immunostaining for type II collagen for each specimen was given a

score of 0-4 for both intensity and extent of staining. For intensity of staining, a score of

0 indicated no staining while a score of 4 indicated the most intense staining. For extent

of staining, a score of 0 indicated no staining, a score of 1 indicated that staining was

noted in 1-25% of the specimen, a score of 2 indicated that staining was noted in 26-50%

of the specimen, a score of 3 indicated that staining was noted in 51-75% of the

specimen, and a score of 4 indicated that staining was noted in 76-100% of the specimen.

The average of scores from two observers was calculated and expressed as means.

Statistical Analysis

Data for each group are represented as median and range. Histologic scores and size values were compared between groups using the Mann-Whitney test (day 17) and

Kruskal-Wallis test with Dunn’s multiple comparisons (day 22). Data analysis was

 performed using the statistical program GraphPad Prism 4 (GraphPad Software, San

Diego, CA). Statistical significance was defined as p < 0.05.

aBone marrow-derived mesenchymal stem cells from male Lewis rats were generously provided by Dr.

Darwin J. Prockop, Margaret R. Wolfe, MS, and Roxanne L. Reger, MS of the Tulane Center for Gene

Therapy (Tulane University Health Sciences Center, New Orleans, LA, USA). These materials were provided by through a grant from NCRR of the NIH, Grant No. P40RR017447. bAdeno-X Rapid Titer Kit, BD Biosciences Clontech, Palo Alto, CA c Histology/Immunohistochemistry Core, Department of Veterinary Biosciences, College of Veterinary

Medicine, The Ohio State University dPathology Core Facility, Department of Pathology, College of Medicine and Public Health, The Ohio State

University

Target Forward primer Reverse primer Probe Aggrecan 5’CCGCTGGTCAGATGGAC 5’GAAGAAGTTGTCGGGCTGGTT3’ 5’TGCATTCGGGCTCAACCTG ACT3’ AA3’ Type 1 5’CTGTGATTTCTCTACTG 5’CCAGTTCTTGGCTGGGATGT3’ 5’TGCATTCGGGCTCAACCTG collagen GCGAAAC3’ AA3’ (α2 chain) (col1a2) *Primer/probe sequences for col2a1 and hbmp2 are proprietary. These primer/probe sets were purchased through the Assays on Demand service of Applied Biosystems (Foster City, CA).

 Table 4.1: Custom-designed primers and probes* for Taqman real-time RT-PCR.

EMPTY ALG No Ad Cell- AdBMP2 Cell- AdBMP6 ALG ALG Cell-ALG RT-PCR: Induction fold change in gene expression*, day 17 for: hbmp2 N/A 1 11.0 N/A col2a1 N/A Not detected 288 N/A aggr N/A N/A 5190 N/A col1a2 N/A N/A 740 N/A Construct size, sq mm, 1.2 (0.96- 15.60 (7.65- N/A N/A day 17 (median, range) 1.751.7)a 24.00)b Construct size, sq mm, N/A 15.60A (6.51- 12.40 (8.75- 4.00 (2.70- day 22 (median, range) 24.00) 16.00)A 6.00)B Cell density scores, day 22 N/A 3 (2.5-3) A 3 (2-4) A 4 (4-4)C (median, range) Cell distribution scores, N/A 3.5 (3-4) 4 (3-4) 2.5 (2-4) day 22 (median, range) Toluidine blue scores, day N/A 3 (0-4) 3 (3-4) 4 (1-4) 22 (median, range) Von Kossa scores, day 22 N/A 1 (1-1) 0 (0-3) 2 (0-2) (median, range)

-∆∆C *Calculated using the 2 T method of Livak and Schmittgen [45] ap = 0.05 vs. No Ad Cell-ALG Bp < 0.05 vs. No Ad Cell-ALG and AdBMP2 Cell-ALG Cp < 0.02 vs. No Ad Cell-ALG and AdBMP2 Cell-ALG

Table 4.2: Quantitative parameters used to evaluate cell-alginate and empty alginate constructs.

∗

Figure 4.1: Median size of cell-alginate constructs from each treatment group. Bars represent median values; error bars demonstrate range. Different letters represent significant differences between groups (p < 0.05). The EMPTY ALG group was used as a negative control to evaluate the effect of the alginate matrix itself. aSignificantly greater than bEMPTY ALG group (p < 0.05) *Significantly greater than AdBMP6 cell-ALG group (p < 0.05), but not significantly different from AdBMP2 cell-ALG group.

400). 400). X Cell-alginate Constructs at day 22, with hematoxylin and eosin, toluidine blue, and von von and blue, toluidine eosin, and hematoxylin with 22, day at Constructs Cell-alginate Figure 4.2: Figure ( staining Kossa

100

400). Mouse Mouse 400). X Immunostaining of ell-alginate constructs for type II collagen at day 22 ( day at collagen II type for constructs ell-alginate of Immunostaining IgG was used as an isotype control (top panel. panel. (top control isotype an as used was IgG Figure 4.3: Figure

101

CHAPTER 5

MESENCHYMAL STEM CELL-MEDIATED GENE DELIVERY OF BONE MORPHOGENETIC PROTEIN-2 IN AN ARTICULAR FRACTURE MODEL

5.1 Summary

In vivo models are valuable to evaluate fracture healing methods. A weight- bearing, distal femoral intercondylar articular osteotomy model was created in the nude rat. Osteotomies were treated with bone marrow-derived mesenchymal stem cells

(BMDMSC), either wild-type (NoAd) or transduced with an adenoviral-bone morphogenetic protein 2 transgene construct (AdBMP2). Cells were delivered in alginate

(ALG) or injected in saline. Controls were empty ALG, saline injections, direct AdBMP2 injection, and untreated osteotomies. Healing was compared using quantitative micro- computed tomography, fluorescent labeling, and histology. At day 14, osteotomy gap area and volume in the AdBMP2 ALG group was significantly greater than any other group (P < 0.003). The group treated with AdBMP2 transduced cells injected in saline

(AdBMP2 cells) had healed with less osteotomy gap area (P < 0.009) and volume (P =

0.01) than untreated controls. In ALG groups, bone healing was impeded by the development of a chondroid mass most pronounced in the AdBMP2 ALG group.

102

Injection of AdBMP2-transduced BMDMSC in saline accelerated bone healing and reconstituted the articular cartilage surface in this distal femoral osteotomy model of articular fracture healing.

5.2 Introduction

It is estimated that 25% of 400,000 human patients undergoing open reduction and internal fixation (ORIF) of fractures each year in the United States will suffer articular fractures resulting in post-traumatic osteoarthritis [1]. Damage to the articular cartilage and chondrocyte death are proposed explanations for the poor outcome associated with many articular fractures [2-6]. While animal models of diaphyseal fractures are well established [7-11], and models involving osteochondral fragments and fissures for the study of cartilage healing have been published [12-14], clinically relevant in vivo models of healing articular fractures are not well described.

Gene delivery to chondrocytes and bone marrow-derived mesenchymal stem cells

(BMDMSC) has been proposed as a mechanism to allow for the augmentation of tissue- engineered bone and cartilage replacement systems. In particular, BMDMSC are attractive targets for gene delivery systems, as they can differentiate, in response to various molecular signals, into chondrocytes and osteoblasts, among other mesenchymal lineages [15-18]. In addition, these cells have the advantage of being readily available for autologous harvest via minimally invasive approaches, and can be expanded in vitro.

Others have demonstrated the ability to induce differentiation of BMDMSC into constructs containing both osteoblasts and chondrocytes. However, to date this directed

103

chondro-osseous differentiation of BMDMSC has only been accomplished when the cells were maintained in chondrogenic and osteogenic differentiating media and/or treated with chondrogenic and osteogenic growth factors in the form of recombinant proteins

[18-21]. Delivery of cDNAs encoding various growth factors, most notably bone morphogenetic protein 2 (BMP2) [18, 22-26], to mesenchymal cells has been described as a method of controlled induction of bone regeneration.

Studies in our laboratory were the first to demonstrate that osteogenic differentiation of BMDMSC in vitro and in alginate could be accelerated with BMP2 gene delivery [27]. Lee and colleagues showed that chondrogenesis of BMDMSC from rabbits could be induced without the use of transforming growth factor (TGF)-β as a chondrogenic differentiating agent if low-intensity ultrasound was used to treat rabbit

BMDMSC-alginate constructs in vitro.a Buxton and colleagues have recently shown that while maintenance of human BMDMSC in three-dimensional polyethylene oxide constructs in a chondrogenic medium containing TGF-β was necessary to induce chondrogenic differentiation, pre-treatment of these cells in monolayer prior to suspension actually disrupted chondrogenic differentiation.b

Bone marrow-derived mesenchymal stem cells have been investigated as components of biologic repair systems in many applications, such as myocardial repair following intramyocardial [28-30] or intracoronary [31] injection of cells and prevention of bone loss in a model of estrogen deficiency using intraperitoneal injection of cells transduced with retroviral RANK-Fc [32]. Intravenously injected syngeneic BMDMSC home to glomeruli in a mesangiolysis model of nephropathy [33], and can improve

104

tendon repair when implanted into tendon defects [34,35]. An advantage of BMDMSC in

bone healing applications is that these cells will localize to the site of a fracture callus

after intravenous injection [36].

The differentiation of BMDMSC in alginate has been studied both in vitro [27,37-

42] and in vivo [43-46]. Chondrogenic and/or osteogenic differentiating media containing growth factors were used to induce the desired differentiation. In our recent report osteogenic differentiation of BMDMSC by human bmp2 and bmp6 could be accelerated in alginate using standard media [27]. Bone morphogenetic protein 2 [37,38,41,42,44-46] and other members of the BMP subfamily of growth factors in the transforming growth factor-β (TGF-β) family of proteins [38-42,45] have repeatedly induced osteogenesis of

BMDMSC via the mechanism of endochondral ossification which has implications for articular fracture healing. Ideally cell-based BMP gene delivery using these BMDMSC in an articular fracture model would use genetically modified cells and the biologic environment to differentiate the BMDMSC into osteoblasts for bone formation and chondroblasts for articular cartilage healing.

The goal of our study was to use a rat model of articular fracture healing to evaluate the effects of BMDMSC-based gene delivery methods on healing of an experimental articular osteotomy of the distal femur. We compared two methods of cell- based gene delivery of hbmp2 for healing of this osteotomy; a three-dimensional alginate construct as a carrier for genetically modified BMDMSC and direct injection of genetically modified BMDMSC suspended in saline solution. Controls included untreated, saline-injection, alginate alone, and direct gene delivery of adenoviral-BMP2

105

construct by percutaneous injection into the osteotomy. Gene expression of BMDMSC within the constructs was confirmed in vivo. Osteotomy gap and bone density of callus were determined by micro-computed tomography (micro-CT) and histology. Our hypotheses were that BMP2-augmented BMDMSC would promote healing of an articular osteotomy and that cells suspended in alginate would be superior to cells suspended in saline solution.

5.3 Results (Note: Methods section appears on page 115.)

In Vitro Imaging of Reporter Gene Product and Histology of Alginate Constructs

Luciferase expression of alginate constructs increased from day 2 to day 4 confirming sustained gene expression and survival of cells. Histology showed cell hypertrophy and presence of extracellular matrix that stained positive with toluidine blue at both time points in both NoAd ALG and AdBMP2 ALG. In the NoAd ALG constructs there was also inconsistent evidence of early mineralization of matrix at day 6. These data provided the rationale for incubating the cell-ALG constructs for 5 days prior to in vivo implantation in the osteotomy sites.

106

In Vivo Imaging of Reporter Gene Product

 The in vivo imaging system (IVIS , Xenogen) detected greater Luc expression

between AdLuc ALG-treated femurs and controls at days 0 (P < 0.01) and 2 (P < 0.05;

Figure 5.1). Values for unoperated femurs were not significantly different from black background (P > 0.05).

Micro-Computed Tomography (Micro-CT)

Two-dimensional micro-CT images were evaluated in the coronal, sagittal, and

axial (transverse) imaging planes (Figure 5.2; Appendix B, Table B1). A software

program (GEHS Microview v.2.0.29, GE Healthcare, Toronto, ONT, Canada) was used

to perform isosurface rendering, resulting in three-dimensional surface reconstructions of

each stifle joint (Figure 5.3; Appendix B, Table B2). A significant difference was found

in osteotomy gap area and gap volume among treatment groups (Figure 5.4; Table 5.1).

The osteotomy gap area and volume in the AdBMP2 ALG group were significantly

greater than in the untreated osteotomy group (P < 0.003), the EMPTY ALG group (P <

0.001), the SALINE group (P < 0.00008), the NoAd cells group (P < 0.0003) and the

AdBMP2 cells group (P < 0.0003; Figure 5.4, Table 5.1). The osteotomy gap area and

volume in the AdBMP2 cells group were significantly smaller than in the untreated

osteotomy group (P < 0.02), the EMPTY ALG group (P < 0.03), the NoAd cells group (P

< 0.02) and the AdBMP2 ALG group (P = 0.000001; Table 5.1).

107

There was no significant difference in mean grayscale value within the osteotomy

gap among groups. There were no significant differences in bone mineral content, bone

mineral density, tissue mineral content, tissue mineral density, voxel values, and bone

volume fraction among groups (Figure 4, Table 1).

Fluorescent Labeling of New Bone Formation and Determination of Bone Porosity

Labeling of bone with calcein was easily identified in the healing lateral femoral

condylar osteotomy of rats when compared with the intact medial femoral condyle.

Uptake of calcein was also identified in the distal femoral and proximal tibial physes

(data not shown).

Mean bone porosity was 47 ± 29% in the SALINE-injected osteotomies versus 74

± 6.9% in the healed osteotomies in the AdBMP2 cells group (P = 0.17). Mean bone porosity of control medial femoral condyles and control proximal tibiae were 58.9 ±

11.2% suggesting increased bone density of healed condyles (sclerosis). Percent bone

porosity of the unaffected condyle was not different among treatment groups nor among

anatomic sites evaluated (P > 0.05). Mean calcein uptake, expressed as percent surface

labeled, was 39.0 ± 16.5% for the healed osteotomized femoral condyle in the AdBMP2 cells group, which was significantly greater than the 6.2 ± 7.1% labeled in the unoperated

medial condyle of the same rats (P = 0.001).

108

Histologic Analysis

Untreated osteotomies contained disorganized fibrous tissue, as evaluated on decalcified histologic sections 14 days after surgery (Figure 5.5A). Osteotomies in the

SALINE group (Figure 5.5E), the NoAd cells group, and the AdBMP2 cells group were characterized by organizing repair tissue. Osteotomies in the alginate groups showed evidence of chondrogenesis that formed round masses of tissue containing round cells in lacunae (Figures 5.5B, 5.5C, and 5.5D) that stained positive for safranin O, consistent with the presence of proteoglycans. These tissue masses were larger with the greatest number and hypertrophy of these round cells in the AdBMP2 ALG group (Figure 5.5D).

In contrast, at day 14 post-injection in rats in the NoAd cells (Figure 5.5F), the articular surface component of the osteotomy site was composed of a mix of fibrous tissue and fibrocartilage. This tissue was more organized and similar to fibrocartilage at the articular surface than the repair tissue in the SALINE (Figure 5.5E) and NoAd cells (Figure 5.5F) groups. The AdBMP2 cells group demonstrated the best reconstitution of a hyaline-like articular cartilage across the healed osteotomy gap (Figure 5.6B).

Staining of calcified sections with Masson’s trichrome was performed on subsets of femora and revealed fibrous tissue in the osteotomy gap in the SALINE group (Fig.

5.7A), completely healed osteotomies in the AdBMP2 cells group (Fig. 5.7B), and the presence of periarticular new bone formation, both in the medial and lateral joint compartments, in the AdBMP2 DIRECT group (Fig. 5.7C).

109

5.4 Discussion

Our results demonstrated enhanced healing of experimental articular osteotomies of the distal femora of rats following ex vivo BMP2 gene delivery using a mesenchymal stem cell-based gene delivery system. The percutaneous injection of BMDMSC genetically modified prior to transplantation to express BMP2 resulted in superior bone healing and a hyaline cartilage surface. An unexpected finding was that osteotomies treated with the same cells in a three-dimensional alginate carrier impeded bone healing.

This was evident in all groups in which alginate was placed in the osteotomy sites. The presence of the alginate gel resulted in the formation of a round tissue mass which did not ossify in any of the osteotomies. While a more proliferative chondroid response was seen within this tissue mass in the AdBMP2 ALG group, bone did not form at the osteotomy site. These masses stained positive with safranin O consistent with the presence of proteoglycans. The morphology of the cells within this mass was consistent with that of hypertrophied chondrocytes.

These results disprove our hypothesis that alginate would be a favorable biologic carrier system for delivery of genetically modified BMDMSC to the site of an articular osteotomy. This hypothesis was based on in vitro data from our laboratory (Zachos and

Bertone, unpublished results 2005) demonstrating up-regulation of gene expression of aggrecan, type II collagen, and type I collagen in the same BMDMSC following ex vivo transduction with the same AdBMP2 transgene construct following five days in culture in media which contained neither chondrogenic nor osteogenic differentiating agents. This was the time point at which the cell-alginate constructs in the present study were

110

delivered to the osteotomy sites as described. The aforementioned in vitro data

demonstrated that the induction fold change in gene expression of type II collagen was

only approximately one-third that of the fold change in gene expression of type I collagen

at day 17, which was the time at which this gene delivery system was evaluated in vivo in

the current study. The in vitro data suggested that rather than predominantly inducing chondrogenic differentiation of BMDMSC, as has been reported by others in in vitro

studies [37-43], a process of chondro-osseous differentiation of BMDMSC was taking

place when these cells are suspended in alginate. The gene expression patterns, as well as

the histologic characteristics, of these cell-alginate constructs, are consistent with the

process of endochondral ossification as it occurs in vivo. The role of BMP2 gene delivery in inducing bone formation via this mechanism in vivo has been well described by others

[18, 44, 46-47]. Because BMP gene delivery may have different effects in vivo [48],

where the potentially synergistic downstream effects of multiple growth factors will play

a role in differentiation of multipotential cells, our hypothesis was that the effects of

BMP2 gene expression on BMDMSC in vivo in diaphyseal bone may differ from effects

on the same cells in a subchondral and metaphyseal bone environment.

Zilberman et al. [44] have shown that BMDMSC genetically engineered to over

express BMP2, when suspended in alginate, formed a cartilage mass, containing

trabecular bone, in a mid-diaphyseal defect of the radius in an immunocompetent mouse

model. Turgeman et al. [11] found similar results in the same osteotomy model using

collagen gels rather than alginate. The predominant tissue of the masses described by

these authors is similar in histologic appearance to the masses of tissue seen in the

111

osteotomy gaps in the AdBMP2 ALG group in the study reported here. This effect of the

alginate to sustain chondrogenesis of BMDMSC over expressing BMP2 obfuscates its

use as a carrier system for these genetically engineered cells when fracture healing is

sought.

Our study is a first to report an exuberant bone proliferation response to in vivo

intra-articular adenoviral BMP2 gene delivery. Within 48 hours of the injection, a mass

was palpable at the joint, resulting in ankylosis within 72 hours. Woven or spongy bone

proliferated from the osteotomy site, and at the site of injection of the AdBMP2 vector.

Massive extensive proliferation of bone was seen extending from this site into the joint

space and periarticular soft tissues in all rats in this group. Although it has been a concern

that excessive bone formation can occur after direct adenoviral hbmp2 gene delivery, this has not been reported to the authors’ knowledge. Several studies have directly injected

AdBMP2 into osteotomies and bone gaps [49-53] without noting this effect. Contributing factors may be the immunoincompetent status of these rats. Heterotopic bone following intramuscular injection in immunoincompetent mice is well known and yet intramuscular injection of Ad vector in immunocompetent rabbits does not consistently produce this effect. [49] Also, the volume of injection (100 µl) may have resulted in leakage of the preparation into the soft tissues around the osteotomy. An argument against muscle forming this heterotopic bone in response to hbmp2 is the general lack of musculature around the stifle and the distinctively periarticular origin of the new bone (Figure 5.6C).

In light of the aforementioned in vitro data published by others, the formation of masses of cartilage in osteotomy sites where alginate was implanted is not a completely

112

unexpected finding. However, we anticipated that in the in vivo environment of the epiphyseal and metaphyseal osteotomy sites, the presence of other osteogenic growth factors would induce osteogenic differentiation of the transplanted cells following the process of endochondral ossification. The fact that the alginate gels, when implanted into the osteotomy sites, actually impeded bone healing was unexpected.

The superior bone healing and remodeling resulting from injection of AdBMP2- transduced BMDMSC into the osteotomy sites three days following creation of the osteotomies was expected. A three day delay in injection was based on previous work from our group demonstrating that immediate direct BMP gene delivery into a rabbit ulnar osteotomy would result in migration of the injected transgene construct away from the osteotomy site if delivered in a saline carrier [49]. The need for an appropriate carrier for the delivery of recombinant BMPs has been clearly demonstrated [54-56]. While this is the result of the rapid diffusion from sites of local administration of the recombinant protein, the use of a carrier in cell-based systems would appear to have the advantage of retaining the cells genetically engineered to up-regulate production of the protein at the anatomic site of interest, and may contribute to osteogenic or chondrogenic differentiation of cells at that site. While this may be the case with other carrier and/or cell types, our study demonstrated that a carrier is not needed in order to incite biologic repair and remodeling of bone.

The results of this study also confirmed a protocol for the use of an in vivo imaging system for the monitoring of luciferase reporter gene expression following the implantation of AdLuc-transduced BMDMSC in alginate beads into osteotomy sites. In

113

vivo molecular imaging systems have not been commonly used to evaluate bone healing,

as this system was developed primarily for the imaging of experimentally created tumors.

A similar system has been used to document reporter gene expression during bone

healing in rodents in vivo. However, this was in a radial diaphyseal defect model [57] and

in calvarial defects and intramuscular injections of the construct of interest [58]. Our

study is the first to report the use of an in vivo imaging system to document and quantify

reporter gene expression in an in vivo model of articular fracture healing. Validation of

this model with this imaging system may be valuable in the further development of in

vivo models of articular fracture healing.

Micro-computed tomographic analysis was a valuable tool for the quantitative

analysis of both mineralized and non-mineralized tissues within the osteotomy gaps, and

for subsequent comparisons to be made among treatment groups. The micro-CT

permitted the identification and quantification of the formation of chondroid masses

which impaired bone healing in the AdBMP2 ALG group. Osteotomy gap volume also

differed significantly among groups, with the AdBMP2 ALG group having the greatest

mean volume and the AdBMP2 cells group having the smallest mean volume.

The results of our study demonstrated that the delivery of BMDMSC to an

articular osteotomy site using an alginate carrier induced the formation of a chondroid

mass which impeded bone healing. In contrast, we found that the injection of BMDMSC

genetically modified ex vivo to express BMP2 can be safely and easily delivered, via a minimally invasive technique, to a distal femoral articular osteotomy site and accelerate bone healing. This clinically relevant in vivo model of articular fracture healing may be

114

used to evaluate the efficacy of other gene delivery systems in healing bone as well as

cartilage and osteochondral defects. Further studies evaluating the utility of this and

similar cell-based gene delivery systems should be conducted in immunocompetent

animals. Ultimately, this model may be used in the development of patient-specific cell-

based gene therapy modalities for delayed union and nonunion fracture healing in both

diaphyseal and articular fractures. Direct (in vivo) delivery of adenoviral-BMP transgene

constructs to articular osteotomy sites is not currently recommended, based on the

vigorous, proliferative bone formation response seen following the use of this method in

our study.

5.5 Methods

Adenoviral vector production

Recombinant adenoviral vectors containing either a 1547 base-pair open reading

frame segment of human BMP-2 or recombinant firefly luciferase (Luc), under the control of the cytomegalovirus promoter, were generated [49,59]. Expression of transgenes was verified in cell culture.

Three-dimensional culture system

Bone marrow-derived mesenchymal stem cells (BMDMSC) from male Lewis rats

were expanded in monolayer culture under standard culture conditions in Dulbecco’s

Modified Eagle Medium (DMEM; Gibco, Grand Island, NY) supplemented with 20%

fetal bovine serum, sodium penicillin at a concentration of 100 units/ml, streptomycin at

115

a concentration of 100 units/ml, and L-glutamine at a concentration of 58.4 mg/ml

(supplemented DMEM). When monolayer cultures had reached approximately 70%

confluence, transduction of cells with the adenoviral (Ad) construct containing the BMP2

transgene cassette (AdBMP2) or AdLuc, described above [49,59], was performed at a

multiplicity of infection of 1350 at 37°C for a transduction time of two hours. The Ad

preparations were then removed from the monolayers and replaced with supplemented

DMEM. Cells were allowed to incubate overnight to achieve expression of the transgene

product.

Monolayers were rinsed with Hank’s balanced salt solution (Gibco) and treated

with 0.25% trypsin-EDTA (Gibco) for five minutes at 37°C. Supplemented DMEM was

added to the monolayers to terminate the trypsin reaction. Cells were subjected to

centrifugation at 450 x g rpm for 10 minutes at room temperature and resuspended in a

1.2% solution of sterile sodium alginate (PRONOVA SLG100, NovaMatrix, FMC

Biopolymer, Oslo, Norway) at a concentration of 5 x 106 cells per 50 µl volume. Aliquots

of the cell-alginate suspension in 50 µl volume were placed into a 102 mM solution of

CaCl2. Constructs were solidified in the CaCl2 solution for 10 minutes, rinsed three times

with sterile 0.9% NaCl, and each 50 µl construct was placed in one well of a 96-well

tissue culture plate (Discovery Labware, BD Biosciences, Bedford, MA) containing

supplemented DMEM and incubated at 37°C with 5% CO2 for 5 days until the time of

surgery.

116

Animals

Forty-six female NIH rnu athymic nude rats (10-12 weeks of age; Charles River

Laboratories, Inc., Wilmington, MA) were used in this study. This study was approved by

the Institutional Laboratory Animal Care and Use Committee (ILACUC) at The Ohio

State University.

Experimental design

Rats undergoing distal femoral osteotomy were placed into one of 9 groups with

regard to treatment of the osteotomy: 4 alginate (ALG) groups, 4 direct injection groups,

and 1 untreated control group. Fifty-microliter (µl) alginate beads were placed in the osteotomy site at the time of creation of the osteotomy in the ALG groups. The ALG groups were: 1) an empty ALG bead (EMPTY ALG; n = 6), 2) untransduced BMDMSC

(NoAd ALG; n = 6), 3) AdLuc-transduced cells (AdLuc ALG; n = 3, used to confirm gene delivery), and 4) AdBMP2-transduced cells (AdBMP2 ALG; n = 7). The direction injection groups were injected 3 days after surgery with 5 x 106 cells in a 100 µl volume

of saline solution (Gey’s Balanced Saline Solution, Gibco). The 4 direct injection groups

were: 1) saline alone (SALINE; n = 6), 2) untransduced BMDMSC in saline (NoAd cells;

n = 6), 3) AdBMP2-transduced BMDMSC in saline (AdBMP2 cells; n = 6), and 4)

AdBMP2 (2 x 1011 particles) in 100 µl of saline (AdBMP2 DIRECT; n = 3).

Additionally, an untreated osteotomy group served as a control (Untreated; n = 6).

117

Articular fracture model

Rats were pre-medicated with subcutaneous injections of meloxicam (Metacam®,

Boehringer Ingelheim Vetmedica, Inc. St. Joseph, MO, 5.0 mg/kg) and butorphanol

tartrate (Torbugesic®, Wyeth, Madison, NJ, 2 mg/kg) for pre-emptive analgesia. General

anesthesia was induced in an anesthesia induction tank using 3% isoflurane in 100%

oxygen and maintained by mask. After aseptic preparation, a volume of 300 µl of 0.1%

lidocaine was injected parallel to the proposed incision site to provide local pre-emptive

analgesia. Surgery was performed on one hind limb with the assistance of 2.5X surgical

magnifying loupes. A lateral intercondylar osteotomy (Figure 5.8) was performed via

lateral parapatellar arthrotomy of the stifle joint. After patella luxation, osteotomy of the

femoral condyle using a No. 11 scalpel blade was initiated at the articular cartilage of the

intercondylar notch and progressed proximally to the level of the lateral distal femoral

physis to exit the cortex just distal to the proximal attachment of the collateral ligament.

Importantly, the collateral ligament was left intact. The osteotomy gap was distracted

using the tip of the blade to ensure complete bone fracture of the condyle (Figure 5.9).

The patella was reduced and the joint capsule was closed with 6-0 polypropylene sutures

 (Surgilene , Davis & Geck, Danbury, CT) in a simple interrupted pattern. A continuous

 subcuticular suture line was placed using 6-0 polypropylene suture (Surgilene , Davis &

Geck). All animals were allowed to ambulate freely in their cages immediately postoperatively. A continuous subcuticular suture line was placed using 6-0 polypropylene

 suture (Surgilene , Davis & Geck) and was augmented using cyanoacrylate tissue

adhesive to appose the skin edges. Immediately following closure of the skin incision, a

118

volume of 250 µl of 0.25% bupivicaine was injected along the incision line to provide

local analgesia with an extended duration. All animals were allowed to ambulate freely in

 their cages immediately post-operatively. Butorphanol tartrate (Torbugesic , 2 mg/kg)

was given subcutaneously every 4 hours for 24 hours for post-operative analgesia.

Delivery of genetically modified cells in a three-dimensional matrix

In 19 rats, cell-matrix constructs (n = 13 rats), or empty matrices (n = 6 rats) were

implanted into the osteotomy site in the distal femur. Wild-type (NoAd) or AdBMP2-

transduced BMDMSC from male Lewis rats (Tulane Center for Gene Therapy) were

suspended in 1.2% sodium alginate, as described previously. Cell-alginate constructs

were then maintained in 37°C 5% CO2 culture conditions in supplemented DMEM in 96-

well tissue culture plates for 5 days following creation of the cell-alginate constructs,a demonstrating induction of endochondral ossification. One 50 µl cell-alginate construct was placed into each distracted osteotomy site using a No. 0 bone curette and a 22-gauge needle to manipulate the construct.

In vitro imaging of reporter gene product and histology

Prior to placement of AdLuc ALG constructs into osteotomy sites in vivo, luciferase transgene product expression was confirmed in vitro in both monolayer

BMDMSC and cells suspended in the three-dimensional alginate constructs (Figs. 10A

 and B) to be implanted surgically, using an in vivo imaging system (IVIS , Xenogen

119

Corporation, Alameda, CA). At days 4 and 6, NoAd ALG and AdBMP2 ALG constructs

were fixed in 10% formalin, sectioned at 6 µm, and stained with toluidine blue and von

Kossa.

In vivo imaging of reporter gene product

Imaging of both operated an unoperated distal femora was performed in vivo to

 quantify duration and intensity of gene (luciferase) expression using IVIS (Xenogen) at

days 0 (immediately post-operatively, Fig. 11), 2, 4, and 6 post-operatively.

Quantification of transgene expression was accomplished by comparing values for counts

(numerical photon data shown with a pseudocolor display

[http://www.xenogen.com/wt/page/software]) and flux (photons of light emitted per

second per square centimeter per steradian), as acquired by a charged cooled coupled

device (CCCD) camera, in the operated stifle to the unoperated contralateral stifle, and

between the operated (AdLuc ALG) and unoperated control stifles and a region of

 interest evaluated as a black background control. Imaging using the IVIS system was

performed by intraperitoneal injection of 150 mg/kg of luciferin 5-15 minutes prior to

imaging under general anesthesia. Previous data from our laboratory have demonstrated

luciferase expression by these Lewis rat BMDMSC in ALG for at least 22 days in vitro

(Zachos and Bertone, unpublished data 2005).

120

Delivery of genetically modified cells in a three-dimensional matrix

Cell-alginate constructs were maintained in 37°C 5% CO2 culture conditions in

supplemented DMEM in 96-well tissue culture plates for 5 days following transduction

before implantation in vivo at surgery. One 50 µl cell-alginate construct was placed into each distracted osteotomy site using a No. 0 bone curette and a 22-gauge needle to manipulate the construct.

Preparation and injection of cell suspensions

Bone marrow-derived mesenchymal stem cells in monolayer (transduced and untransduced) were detached with trypsin-EDTA (Gibco), subjected to centrifugation at

450 x g for 10 minutes at room temperature. Cells were then suspended in saline (Gey’s

Balanced Salt Solution, Gibco) at a concentration of 5 x 107 cells/ml, and 100µl injected

into the osteotomy site of each anesthetized rat (5 x 106 cells/100 µl). To accomplish this,

rats were anesthetized with isoflurane, as previously described. A suspension containing

5 x 106 BMDMSC in 100 µl saline was aspirated from a 1.5 ml microcentrifuge tube into

a 1.0 ml syringe with either a 20-gauge or an 18-gauge needle attached. The stifle (knee)

joint was flexed to 45° and the needle was introduced into the lateral parapatellar

compartment of the stifle joint. The needle was advanced proximolaterally until the bevel

of the needle contacted the lateral femoral condyle. The needle was then directed further

proximally and laterally, while making contact with the bone, until the needle entered the

121

osteotomy site. With the needle held firmly in the osteotomy gap, the plunger was depressed, depositing 5.0 x 106 BMDMSC, either wild-type (NoAd) or genetically modified to express BMP2 into the osteotomy site.

Micro-computed tomography (micro-CT)

Forty-six stifle joints were scanned in air using a MS-130 high resolution micro-

CT scanner (GE Healthcare, Toronto, ONT, Canada) at 28 µm voxel resolution, at 75 kV and 75 mA. Scans were performed in the Scaffold Engineering Laboratory in the

Department of Biomedical Engineering at the University of Michigan (Ann Arbor, MI).

Isosurface rendering was performed to obtain images reconstructed in three dimensions

(Figure 5.3). The bone mineral density (BMD, mg/cc) and area (mm2) of each specimen was calculated by defining a region of interest completely contained within the osteotomy gap and an appropriate threshold level for bone using GEHC Microview software

(v.2.0.29, GE Healthcare, Toronto, ONT, Canada). Two-dimensional micro-CT images in the coronal plane in JPEG format were created using GEMS Microview and imported into the ImageJ software program (Rasband, WS, ImageJ, U.S. National Institutes of

Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997-2005). This software was used to trace the exact outline of the osteotomy gap in each specimen and to calculate total osteotomy gap area and volume from two-dimensional images in the axial

(transverse), coronal, and sagittal computed tomographic imaging planes. A region of interest was selected entirely within the osteotomy gap (if present) in each femur, and mean grayscale value, bone volume of the same selected region within the osteotomy

122

gap, voxel values of the region, bone mineral content, bone mineral density, tissue mineral content, tissue mineral density, and bone volume fraction of the gap were calculated (GEHC Microview).

Fluorescent labeling of new bone formation and determination of bone porosity

Nine rats were injected with a 2% solution of calcein (Sigma P/N C-0875, Sigma

Chemical Company, St. Louis, MO) at a dose of 20 mg/kg/rat given subcutaneously on post-operative day 7 and as an intraperitoneal injection (diluted 1:10 in 0.9% NaCl) on post-operative day 11 [60-62]. These rats were representative of the SALINE [n = 3],

AdBMP2 cells [n = 3], or AdBMP2 DIRECT [n = 3]) groups. Sections of calcified bone were prepared using EXAKT cutting and grinding systems (EXAKT Technologies, Inc.,

Oklahoma City, OK) and were examined using fluorescence microscopy. A grid was used to determine relative uptake of calcein by anatomic region. In 1 to 3 fields of view for each of three anatomic locations (lateral femoral condyle, proximal tibia, and medial femoral condyle), 81 points were evaluated. At each point, the presence or absence of bone and the presence or absence of fluorescence was noted. These values were used to calculate percent bone porosity and percent surface labeled, respectively.

Statistical Analysis

Values for percent bone porosity and percent surface labeled were compared between groups using the Kruskal-Wallis test with Dunn’s post-test for nonparametric

 multiple comparisons (GraphPad PRISM 4, GraphPad Software, San Diego, CA).

123

Values for osteotomy gap area and volume were compared between groups using two- way analysis of variance with least squares difference post-test (STATISTICA, StatSoft,

Inc., Tulsa, OK). Factor 1 was the presence or absence of alginate. Factor 2 was treatment, defined as: untreated, vehicle (alginate or saline), cells in vehicle, or BMP2- transduced cells in vehicle. A P-value of < 0.05 was considered significant. Luciferase

 expression data obtained using the IVIS system were compared between treatment

 groups using the Mann-Whitney U test, (GraphPad PRISM 4). Values for percent bone porosity and percent surface labeled were compared among groups using the Kruskal-

 Wallis test and Dunn’s post-test (GraphPad PRISM 4). A P-value of < 0.05 was considered significant.

Histologic analysis

Rats were euthanized by CO2 inhalation followed by cervical dislocation on day

14 after treatment. Osteotomized limbs were harvested immediately following euthanasia. Limbs were harvested and fixed in 10% neutral buffered formalin (NBF) for

72 hours at room temperature. The limbs were decalcified in 10% ethylenediaminetetraacetic acid (EDTA, pH 7.4) for 7 days. (The EDTA was changed after 72 hours during decalcification.) The limbs were placed in 10% NBF for transport to the histotechnology facility, where they were dehydrated in a series of alcohols, embedded in paraffin, cut into 6-µm sections, stained with hematoxylin and eosin or

124

safranin O and fast green, and examined using light microscopy. Stained sections were reviewed by 3 of the authors (TAZ; ALB; and SEW), including a board-certified pathologist sub-specializing in orthopaedic pathology (SEW). aLee, H.J., Park, S.R., and Min, B.-H. Effects of low intensity ultrasound on chondrogenesis of bone marrow mesenchymal stem cells. Trans Orthop Res Soc 2004, Paper No. 814. bBuxton, A.N., Marchant, R., Watts, K., West, J., Yoo, J.U., and Johnstone, B. In vitro chondrogenesis of mesenchymal progenitor cells in a photopolymerizable poly(ethylene oxide) semi-interpenetrating network. Trans Orthop Res Soc 2006, Paper No. 776.

125

Untreated EMPTY ALG NoAd ALG AdBMP2 ALG SALINE NoAd Cells AdBMP2 Cells Osteotomy gap area 3688 ± 1921a 3189 ± 1720a 2453 ± 1607a,c 6843 ± 1546b 1987 ± 1257a,c 3429 ± 1586a 0 ± 0c (sq pixels) Mean grayscale 1269 ± 321.8 1159 ± 357.3 1264.8 ± 202.3 702.2 ± 165.3 1470.7 ± 313.1 1137 ± 308.4 1413 ± value (bytes) 260.9 Total volume (mm3) 0.46 ± 0.30a 0.12 ± 0.06a 0.49 ± 0.34a 0.89 ± 0.24b 0.26 ± 0.19a 0.34 ± 0.17a,c 0 ± 0c Bone volume (mm3) 0.40 ± 0.14 0.39 ± 0.25 0.51 ± 0.13 0.26 ± 0.12 0.22 ± 0.06 0.18 ± 0.06 0.37 ± 0.10 4 4 4 4 4 4 4 Voxels (Hounsfield 3.77 x 10 ± 1.01 x 2.59 x 10 ± 4.44 x 10 ± 4.94 x 10 ± 2.19 x 10 ± 2.41 x 10 ± 2.52 x 10 ± 4 4 4 3 3 3 3 Units) 10 1.12 x 10 1.11 x 10 8.65 x 10 6.21 x 10 4.99 x 10 5.81 x 10 Bone mineral 0.26 ± 0.07 0.23 ± 0.14 0.32 ± 0.08 0.19 ± 0.06 0.15 ± 0.02 0.12 ± 0.02 0.21 ± 0.06 content (mg) Bone mineral 317.4 ± 80.26 290.9 ± 89.75 329.9 ± 54.93 364.8 ± 66.28 374.6 ± 77.81 287.5 ± 78.12 364.8 ± density (mg/cm3) 66.28 126 Tissue mineral 0.21 ± 0.08 0.21 ± 0.14 0.26 ± 0.07 0.11 ± 0.06 0.11 ± 0.03 0.09 ± 0.03 0.19 ± 0.05 content (mg) Tissue mineral 464.78 ± 37.42 454.2 ± 37.63 509.2 ± 24.58 408.9 ± 19.48 488.5 ± 26.92 449.3 ± 30.22 473.2 ± density (mg/cm3) 19.60 Bone volume 0.49 ± 0.17 0.45 ± 0.18 0.51 ± 0.11 0.23 ± 0.08 0.62 ± 0.18 0.45 ± 0.16 0.63 ± 0.14 fraction

Table 5.1: Parameters measured using micro-computed tomography (micro-CT) images. Values are given as mean ± standard error of the mean. Different letters within a row represent different levels of significance. bP < 0.02 cP < 0.015 Untreated = untreated osteotomy; EMPTY ALG = 50 µl alginate bead containing no cells; AdBMP2 ALG = 50 µl alginate bead containing 5 x 106 bone marrow-derived mesenchymal stem cells transduced with an adenoviral-bone morphogenetic protein 2 transgene construct; NoAd ALG =50 µl alginate bead containing 5 x 106 untransduced bone marrow-derived mesenchymal stem cells; SALINE injection = 100 µl injection of saline containing no cells; AdBMP2 cells = injection of 5 x 106 bone marrow-derived mesenchymal stem cells transduced with an adenoviral-bone morphogenetic protein 2 transgene construct suspended in 100 µl injection of saline; NoAd cells = injection of 5 x 106 untransduced bone marrow-derived mesenchymal stem cells suspended in 100 µl injection of saline

Luciferase Expression: Rat Model 1.00E+05 * 9.00E+04

s 8.00E+04 7.00E+04 6.00E+04 5.00E+04 Counts 4.00E+04 g 3.00E+04 2.00E+04 1.00E+04 0.00E+00 Operated Femur Unoperated Black Day 0 (AdLuc ALG) Control Femur Background Day 2 Day 4 Region Evaluated

Figure 5.1: Expression of luciferase, quantified as counts (numerical photon data count shown with a pseudocolor display [http://www.xenogen.com/wt/page/software]), in femora of rats receiving alginate constructs with cells transduced with adenoviral luciferase gene (AdLuc) into an articular osteotomy. Expression in operated femora was significantly greater than that in unoperated contralateral control femora and background. *p < 0.01, gp < 0.05

127

Figure 5.2: Two-dimensional micro-computed tomographic (micro-CT) images (GEHC Microview v.2.0.29, GE Healthcare, Toronto, ONT, Canada) in the coronal imaging planes demonstrating appearance of osteotomy gap (arrows) among treatment groups. The AdBMP2 ALG group (C) had greater gap (P < 0.05) and the AdBMP2 cells group (F) had no gap (P < 0.0001), and the AdBMP2 DIRECT group (H) had periarticular bone formation (arrowheads). In the Untreated (a), EMPTY ALG (b), and NoAd ALG (c) groups, the osteotomy gap is highlighted in red, adjacent bone is highlighted in green, and menisci are highlighted in yellow. 128

Figure 5.2 (continued): In the AdBMP2 ALG (d), SALINE (e), and NoAd cells (f) groups, the osteotomy gap is highlighted in red, adjacent bone is highlighted in green, and menisci are highlighted in yellow.

129

Figure 5.2 (continued): In the AdBMP2 cells group (g), the healed, remodeled bone of the distal femur is highlighted in green. In the AdBMP2 DIRECT (h) group, the intact bone of the distal femur is highlighted in green and the periarticular new bone proliferating from the osteotomy site is highlighted in blue.

130

Figure 5.3: Selected isosurface renderings of micro-computed tomographic images demonstrating unhealed bone in the Untreated group (arrows; A, a), healed bone in the AdBMP2 cells group (BMDMSC transduced with AdBMP2 and injected in saline; B, b), periarticular bone (arrowheads) in the AdBMP2 DIRECT (direct gene delivery) group (arrowheads; C, c). In (a), osteotomy gap is highlighted in red adjacent to normal bone (green). In the AdBMP2 cells group (b), the healed distal femur is green. In the AdBMP2 DIRECT group (c) periarticular new bone is highlighted in blue, and menisci are yellow. 131

Figure 5.4: Osteotomy gap area values calculated from micro-computed tomographic (micro-CT) images at day 14. Different letters indicated significant differences between groups. bP < 0.002; cP < 0.0003

132

Figure 5.5: Histology of decalcified sections, stained with safranin O/fast green (X400), representative of each treatment group 14 days after treatment: (A) Untreated, (B) EMPTY ALG, (C) No Ad ALG, (D) AdBMP2 ALG, (E) SALINE, (F) No Ad cells, (G) AdBMP2 cells. (A), (E), and (F) show disorganized fibrous tissue. (B) shows some cells isolated in small lacunae suggesting chondrogenesis. (C) and (D) show marked chondrogenesis with chondrocyte hypertrophy most dramatic in (D) and mixed with fibrous tissue in (C). (G) shows bone bridging of the osteotomy gap.

133

Figure 5.6: Comparison of histologic appearance of articular surfaces. (A) Fibrous tissue fills a defect in the articular surface in an osteotomy site treated with injection of saline (white arrows). There is minimal bone present in the osteotomy gap (black arrow). (B) Restoration of the weight bearing surface of the articular cartilage (white arrows) is seen in the osteotomy treated with an injection of bone marrow-derived mesenchymal stem cells transduced with an adenoviral-bone morphogenetic protein 2 transgene construct. Remodeled bone is present in the osteotomy gap (black arrows). (Both H&E X400)

134

Figure 5.7: Calcified sections: light microscopic appearance of lateral femoral condyles of rats in SALINE Injection (i.e. control) group (A) and the AdBMP2xBMDMSC Cell Injection group (B). Arrows in A indicate the medial and lateral borders of the unhealed osteotomy gap. Completely healed and remodeled bone with an increased density of the trabecular bone is seen in B. (Masson’s trichrome X40)

Figure 5.7 (continued): Brackets in C delineate the extent of periarticular new bone formation laterally in the AdBMP2 direct gene delivery group. The adenoviral-transgene construct was injected into the lateral femoral condylar osteotomy site. (Masson’s trichrome X40)

135

Figure 5.8: A schematic, drawn from radiographs of the rat stifle joint, depicting the lateral intercondylar osteotomy (dotted line) performed in athymic nude rats (images by Tim Vojt).

136

A B

Figure 5.9: Lateral intercondylar osteotomy of the distal femur in a 180-gram nude rat. A Intra-operative view. Distracted osteotomy site is delineated by arrows. Proximal aspect is at the bottom of the image. B Radiograph taken immediately after osteotomy (between arrows).

Figure 5.10: Use of the In Vivo Imaging System (IVIS®) on live bone marrow-derived mesenchymal stem cells (BMDMSC) in vitro genetically modified to express the reporter gene product luciferase. A Monolayer culture showing bioluminescence of luciferase expressing BMDMSC. B Three-dimensional cell-alginate constructs (50 µl each, containing 5 x 106 cells per construct) wells showing bioluminescent (right arrow) cells expressing luciferase and wells (left arrow) are untransduced.

137

Figure 5.11: Use of the In Vivo Imaging System (IVIS®, Xenogen Corporation, Alameda, CA) to detect luciferase reporter gene product expression by genetically modified bone marrow-derived mesenchymal stem cells (BMDMSC) in vivo. A 50 µl alginate bead containing 5 x 106 BMDMSC transduced with an adenoviral luciferase (AdLuc) construct was placed into a distracted lateral femoral condylar osteotomy site in the left distal femur. Luciferase expression, as detected by a charged cooled coupled device (CCCD) camera system, in the left femur (Region Of Interest [ROI] 1) is shown by the display of color and is denoted in counts on the pseudocolor scale on the right. The unoperated contralateral (right) distal femur (ROI 2) and adjacent black background (ROI 3) serve as controls for the measurement of counts.

138

APPENDIX A: MICROARRAY DATA FROM IN VITRO STUDIES

139

Genbank Name Gene Function Fold Change Accession Function d2 AdBMP2 vs. d2 AdBMP6 vs. d2 AdLuc vs. d12 AdBMP2 vs. d12 AdBMP6 vs. Number Classification untransduced untransduced untransduced untransduced untransduced <48ha <48ha <48ha <48ha <48ha AB106115 Smad6 Endochondral Regulation of transcription of 381.1 14.9 -- -- 19.7 osteogenesis BMPs AF510665 Bone Endochondral Embryonic -- 362.0 -- -- 78.8 morphogenetic osteogenesis development/morphogenesis; protein (BMP6) signal transduction precursor BM781374 Fibulin 1 Endochondral Calcium ion binding; ECM 5.3 6.1 -- 8.0 8.6 (FBLN1), osteogenesis constituent transcript variant C BM781120 Retinoic acid Endochondral Negative regulation of cell ------6.1 6.5 receptor osteogenesis proliferation responder (tazarotene induced) 1 (RARRES1)

140 AF034691 Procollagen Endochondral Major fibrillar protein ------8.6 alpha 1 (I) osteogenesis component of connective tissues (COL1A1) AF117954 Procollagen Endochondral Found in extensible connective ------4.9 alpha-1 type III osteogenesis tissues, frequently in association precursor with type I collagen (COL3A1) AY005821 Parathyroid Endochondral Regulates endochondral bone ------4.3 hormone- osteogenesis development related peptide BI960839 Exostoses Endochondral Signal transduction; -- 3.0 ------(multiple) 1 osteogenesis glycosaminoglycan biosynthesis; (EXT1) skeletal development D50326 Inhibin beta A Endochondral Cell growth/maintenance; signal 3.5 -- -- 3.0 4.3 subunit osteogenesis transduction; skeletal development; apoptosis BM780462 PAPS Endochondral Nucleotide and nucleic acid ------4.9 3.7 synthetase-2 osteogenesis metabolism; skeletal (3'- development; sulfate phosphoadenosi assimilation ne 5'- phosphosulfate synthase 2) (PAPSS2) AF547432 Golgi apparatus Endochondral Associated with fibroblast ------18.4 protein osteogenesis growth factor (FGF) binding

AJ319906 Fibroblast Endochondral Signal transduction, cell -- -- -4.6 -- -- growth factor 2 osteogenesis growth/maintenance; skeletal (fgf2) development AB106118 ALK5 for TGF Endochondral Embryonic -- 13.0 18.4 -- -- beta receptor osteogenesis development/morphogenesis type I CD535200 Cdc42 guanine Chondrogenes Signal transduction 3.5 3.2 ------nucleotide is exchange factor (GEF) 9 (ARHGEF9, chondrocyte- derived) AF325902 Cartilage Chondrogenes Noncollagenous extracellular ------55.7 55.7 oligomeric is matrix protein matrix protein (COMP) AB072934 Angiomodulin Chondrogenes Insulin-like growth factor ------3.7 4.3 is binding BM780480 Chemokine (C- Inflammation Cell motility, chemotaxis, ------21.1 7.0

141 X-C motif) immune response/inflammation, ligand 10 positive regulation of cell (CXCL10) proliferation BM734843 Protease Inflammation Inflammatory response; wound ------10.6 14.9 inhibitor 3, healing; negative regulation of skin-derived protease secretion by neutrophils (SKALP) CD468265 Interleukin 8 Inflammation Inflammation, signal -- 3.25 -- -- 9.8 (IL8) transduction U62529 MMP 3 Inflammation Collagen catabolism 3.2 3.5 ------CD535170 Histatin 1 Inflammation Inhibits calcium phosphate ------3.5 (HTN1) precipitation; defense response to bacteria & fungi; ossification M64087 Tumor necrosis Inflammation Regulation of transcription; 3.5 ------factor-alpha signal transduction; anti- (TNF-α) apoptosis; apoptosis; necrosis BI961838 Capping protein Inflammation Barbed-end actin filament -3.03 -3.25 -- -4.0 -4.0 (actin filament), capping; protein complex gelsolin-like assembly; response to pest, (CAPG) pathogen or parasite CD465523 2'-5' Inflammation Inflammation -3.48 ------oligoadenylate synthetase 2 short isoforms CD471347 CD68 antigen Inflammation Transmembrane glycoprotein ------4.3 (CD68) highly expressed by monocytes

and tissue macrophages BM734900 MMP 9 Inflammation Collagen catabolism: degrades ------4.3 collagen types IV and V AJ251189 Monocyte Inflammation Chemotaxis; immune & ------5.0 chemoattractant inflammatory responses; signal protein-1 (mcp- transduction 1) BM780537 Transglutamina Other G-protein-coupled receptor ------13.9 21.1 se 2 (TGM2), protein signaling pathway, transcript peptide cross-linking, positive variant 1 regulation of cell adhesion CD465947 Dual specificity Other Inactivate target kinases by ------18.4 29.9 phosphatase 2 dephosphoryla- (DUSP2) tion; negatively regulates members of MAPK superfamily CD469333 Myeloid cell Other Cellular defense response; ------17.1 10.6 nuclear regulation of transcription differentiation antigen (MNDA)

142 BI961105 PRG1 (p53- Other Anti-apoptosis; apoptosis; cell 6.96 12.13 -- 5.7 4.9 responsive gene growth/maintenance 1) (Aliases: IEX-1, IER-3) BM781319 Cyclin D2 Other Cytokinesis, regulation of cell ------7.5 5.7 (CCND2) cycle BM781147 Inositol 1,4,5- Other Calcium ion transport,signal ------6.5 4.6 triphosphate transduction receptor, type 1 (ITPR1) BI961486 Diphtheria Other Muscle development, signal ------4.9 5.7 toxin receptor transduction (heparin- binding epidermal growth factor- like growth factor) (DTR) BM735056 Interferon Other Regulation of transcription -- 3.0 ------regulatory factor 1 CD471623 Uridine Other Tightly regulates concentration ------4.6 3.7 phosphorylase of uridine, a pyrimidine (UPP2) nucleoside essential for the synthesis of RNA and biomembranes

U31699 Gelsolin@ Other Regulation of cells growth and ------4.9 3.5 apoptosis AF508034 Plasminogen Other Blood coagulation; regulated by ------5.3 8.6 activator cell adhesion** inhibitor-1 (PAI-1) AJ439891 Keratinocyte Other Potent epithelial cell-specific ------3.2 growth factor growth factor, predominant in (fgf-7) keratinocytes but not fibroblasts and endothelial cells; re- epithelialization of wounds BM780543 Glycoprotein Other Coagulation; cell-matrix ------3.0 -- IIIa (ITGB3) adhesion; integrin signaling U95039 Tissue inhibitor Other Inhibitor of MMPs ------3.0 of metalloproteinase-1 (TIMP-1) CD467407 Zinc finger Other Likely functions in regulating ------3.0 protein 36, C3H transcription response to growth type-like 1 factors

143 (ZFP36L1)

BM780597 CCAAT/enhanc Other Regulation of transcription -- -3.7 ------er binding protein (C/EBP), delta (CEBPD) CD535443 HMG-box Other Signal transduction (Wnt -- -3.0 ------transcription signaling); regulation of factor TCF-3 transcription CD465025 Dual specificity Other Cell cycle regulation -- -3.3 ------phosphatase 6 (DUSP6) CD535297 High mobility Other Regulation of transcription; ------group AT-hook protein complex assembly 1 (HMGA1), transcript variant 2 BI961081 Xanthene Other Electron transport ------dehydrogenase (XDH) BM735440 Short form Other Cell growth/maintenance; ------transcription regulation of transcription factor C-MAF AF027335 Prostaglandin Other Prostaglandin synthesis; cyclo- -4.0 ------G/H synthase 2 oxygenase pathway

(Alias: COX2) BM780291 Phosphatidic Other Cell growth and/or maintenance; ------3.3 -4.0 acid germ cell migration; lipid phosphatase metabolism type 2B (PPAP2B) MMP = matrix metalloproteinase; MAPK = mitogen-activated protein kinase aIndicates untransduced cells in monolayer cultures for less than 48 hours @Fujita H. [An actin-regulatory protein, gelsolin, functions as a regulator of cell growth and apoptosis] Japanese. Seikagaku 2002;74: 135-39. **Lee CC, Shyu KG, Lin S, et al. Cell adhesion regulates the plasminogen activator inhibitor-1 gene expression in anchorage-dependent cells. Biochem Biophys Res Commun 2002;291:185-90.

Table A1: Genes differentially regulated (>3-fold change) in response to transduction with Ad-BMP-2, Ad-BMP-6, or Ad-Luc 144

Biological Process of Gene Product Matrix Proteins Signaling Proteins Type I collagen alpha 2 chain Indian hedgehog (IHH) Type II collagen Smoothened Aggrecan core protein Noggin Biglycan Epidermal growth factor (EGF) Decorin Fibroblast growth factor 1 (FGF1) Dermatan sulfate proteoglycan II Fibroblast growth factor 2 (FGF2) Fibronectin Fibroblast growth factor receptor (FGFR) FGF-7 receptor 2IIIb FGF-2 receptor IIIc Smad1 Smad3 Smad4 Smad5 Smad7 Sox9 Core binding factor alpha 1 (CBFA1) subunit (Alias: Runx2) Inhibin, alpha subunit PreproIGF-I PreproIGF-Ia IGF-II TGF-β receptor type II TGF-β1 VEGF receptor flt IGF = insulin-like growth factor; FGF = fibroblast growth factor; COMP = cartilage oligomeric matrix protein; TGF = transforming growth factor; VEGF = vascular endothelial growth factor

Table A2: Genes relevant to osteogenic and/or chondrogenic differentiation whose expression was not altered by adenoviral transduction of hbmp2 or hbmp6 at day 2 or day 12

145

APPENDIX B: SUPPLEMENTARY MICRO-COMPUTED TOMOGRAPHY DATA

146

Figure B1: Representative two-dimensional micro-computed tomographic (micro-CT) images (GEHC Microview v.2.0.29, GE Healthcare, Toronto, ONT, Canada) in the coronal (A, D, G, J, M, P, S, V), sagittal (B, E, H, K, N, Q, T, W), and axial (transverse; C, F, I, L, O, R, U, X) imaging planes demonstrating differences in appearances of osteotomy gap (arrows) among treatment groups.

147

Figure B1. (continued). The AdBMP2 ALG group (J, K, L) had greater gap (P < 0.05) ,the AdBMP2 cells group (S, T, U) had no gap (P < 0.0001), and the AdBMP2 DIRECT group (V, W, X) had periarticular bone formation (arrowheads).

148

Figure B1. (continued).

149

Figure B2: Representative isosurface renderings of micro-computed tomographic (micro-CT) images (GEMS Microview v.2.0.29, GE Healthcare, Toronto, ONT, Canada) demonstrating cranial (A, D, G, J, M, P, S, V), caudal (B, E, H, K, N, Q, T, W), and craniolateral (C, F, I, L, O, R, U, X) unhealed bone in the Untreated (untreated osteotomy; A, B, C), EMPTY ALG (alginate without cells; D, E, F), NoAd ALG (untransduced bone marrow-derived mesenchymal stem cells [BMDMSC] suspended in alginate; G, H, I), AdBMP2 ALG (BMDMSC transduced with AdBMP2 and suspended in alginate; J, K, L), SALINE (saline injection without cells; M, N, O), and NoAd cells (unstransduced BMDMSC injected in saline suspension; P, Q, R), groups. Arrows indicate osteotomy gaps.

150

Figure B2 (continued): Arrows indicate osteotomy gaps.

151

Figure B2 (continued): Healed bone is present in the AdBMP2 cells group (bone marrow-derived mesenchymal stem cells [BMDMSC] transduced with AdBMP2 and injected in saline suspension; S, T, U) and periarticular bone (arrowheads) is present in the AdBMP2 DIRECT (direct AdBMP2 gene delivery) group (V, W, X).

152

BIBLIOGRAPHY

CHAPTER 1 (GROWTH FACTORS: INTERACTIONS AND THERAPEUTIC IMPLICATIONS FOR MUSCULOSKELETAL CONNECTIVE TISSUES)

1. Day MJ, Butterworth SJ, Palmer MR, et al. Characterization of wear debris associated with aseptic loosening of a canine hip. J Comp Pathol 1998;119:89-93.

2. Edwards MR, Egger EL, Schwarz PD. Aseptic loosening of the femoral implant after cemented total hip arthroplasty in dogs: 11 cases in 10 dogs (1991-1995). J Am Vet Med Assoc 1997;211:580-586.

3. Dyce J, Olmstead ML. Removal of infected canine cemented total hip prostheses using a femoral window technique. Vet Surg 2002;31:552-560.

4. Halling KB, Lewis DD, Cross AR, et al. Complication rate and factors affecting outcome of olecranon osteotomies repaired with pin and tension-band fixation in dogs. Can Vet J 2002;43:528-534.

5. Cook JL, Tomlinson LF, Stoll MR, et al. Arthroscopic removal and curettage of osteochondrosis lesions on the lateral and medial trochlear ridges of the talus in two dogs. J Am Anim Hosp Assoc 2001; 37:75-80.

6. Carter BG, Bertone AL, Weisbrode SE, et al. Influence of methylprednisolone acetate on osteochondral healing in exercised tarsocrural joints of horses. Am J Vet Res 1996;57:914-922.

7. Grondahl AM, Jansen JH, Teige J. Accessory ossification centers associated with osteochondral fragments in the extremities of horses. J Comp Pathol 1996;114:385-398.

153

8. Todhunter RJ, Freeman KP, Yeager FE, et al. Effects of exercise on the development of osteoarthritis in equine carpal joints with osteochondral defects. Vet Surg 1993;22:330- 342.

9. Stromberg B, Rejno S. Osteochondrosis in the horse. I. A clinical and radiologicinvestigation of osteochondritis dissecans of the knee and hock joint. Acta Radiol Suppl 1978;378:139-152.

10. Anastasiou A, Skioldebrand E, Ekman S, et al. Ex vivo magnetic resonance imaging of the distal row of equine carpal bones: assessment of bone sclerosis and cartilage damage. Vet Radiol Ultrasound 2003;44:501-512.

11. Gordon WJ, Conzemius MG, Riedesel E, et al. The relationship between limb function and radiographic osteoarthrosis in dogs with stifle osteoarthrosis. Vet Surg 2003;32:451-454.

12. Mortellaro CM. Pathophysiology of osteoarthritis. Vet Res Commun 2003;27 Suppl 1:75-78.

13. Fernihough JK, Innes JF, Billingham ME, et al. Changes in the local regulation of insulin-like growth factors I and II and insulin-like growth factor-binding proteins in osteoarthritis of the canine stifle joint secondary to cruciate ligament rupture. Vet Surg 2003;32:313-323.

14. Edinger DT, Hayashi K., Hongyu Y, et al. Histomorphometric analysis of the proximal portion of the femur in dogs with osteoarthritis. Am J Vet Res 2000;61: 1267- 1272.

15. Kealy RD, Lawler DF, Ballam JM, et al. Evaluation of the effect of limited food consumption on radiographic evidence of osteoarthritis in dogs. J Am Vet Med Assoc 2000;217: 1678-1680.

16. Kurochi K, Cook JL, Kreeger JM. Mechanisms of action and potential uses of hyaluronan in dogs with osteoarthritis. J Am Vet Med Assoc 2002;221: 944-950.

17. Volk SW, Kapatkin AS, Haskins ME, et al. Gelatinase activity in synovial fluid and synovium obtained from healthy and osteoarthritic joints of dogs. Am J Vet Res 2003;64: 1225-1233.

18. MacWilliams PS, Friedrichs KR. Laboratory evaluation and interpretation of synovial fluid. Vet Clin North Am Small Anim Pract 2003; 33:153-178.

19. Breitschwerdt EB, Abrams-Ogg AC, Lappin MR, et al. Molecular evidence supporting Ehrlichia canis-like infection in cats. J Vet Intern Med 2002;16:642-649. 154

20. Ollier WE, Kennedy LJ, Thomson W, et al. Dog MHC alleles containing the human RA shared epitope confer susceptibility to canine rheumatoid arthritis. Immunogenetics 2001;53:669-673.

21. Pfizer Animal Health. Proprietary market research, survey of 200 veterinarians. 1996. Available at: www.rimadyl.com/facts/index.html. Accessed Dec. 1, 2003.

22. Hardie EM, Roe SC, Martin FR. Radiographic evidence of degenerative joint disease in geriatric cats: 100 cases (1994-1997). J Am Vet Med Assoc 2002;220:628-632.

23. Johnson JM, Johnson AL. Cranial cruciate ligament rupture. Pathogenesis, diagnosis, and postoperative rehabilitation. Vet Clin North Am Small Anim Pract 1993;23:717-733.

24. Miyasaka KC, Daniel DM, Stone ML. The incidence of knee ligament injuries in the general population. J Knee Surg 1991;4:3-8.

25. Teitz CC, Garrett WE Jr, Miniaci A, et al. Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Tendon Problems in Athletic Individuals. J Bone Joint Surg Am 1997;46: 569-582.

26. Silver IA, Brown PM, Goodship AE. A clinical and experimental study of tendon injury, healing and treatment in the horse. Equine Vet J Suppl 1983;1:1-43.

27. Dowling BA, Dart AJ, Hodgson DR, et al. The effect of recombinant equine growth hormone on the biomechanical properties of healing superficial digital flexor tendons in horses. Vet Surg 2002;31(4):320-324.

28. Sutter WW, Kaneps, AJ, Bertone AL. Comparison of hematologic values and transforming growth factor-b and insulin-like growth factor concentrations in platelet concentrates obtained by use of buffy coat and apheresis methods from equine blood. Am J Vet Res 2004;65: 924-930.

29. Stiles J, Townsend W, Willis M. Use of a caudal auricular axial pattern flap in three cats and one dog following orbital exenteration. Vet Ophthalmol 2003;6:121-126.

30. Trevor PB, Smith MM, Waldron DR, et al. Clinical evaluation of axial pattern skin flaps in dogs and cats: 19 cases (1981-1990). J Am Vet Med Assoc 1992;201:608-612.

31. Laing EJ. Problems in wound healing associated with chemotherapy and radiation therapy. Probl Vet Med 1990;2:433-441.

32. Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 4th ed. New York: Garland Science, Taylor & Francis Group, 2002;492, 1294.

155

33. Wu D, Razzano P, Grande DA. Gene therapy and tissue engineering in repair of the musculoskeletal system. J Cell Biochem 2003;88:467-481.

34. Crystal RG. Transfer of genes to humans: Early lessons and obstacles to success. Science 1995;270:404-410.

35. Oligino TJ, Yao Q, Ghivizzani SC, et al. Vector systems for gene transfer to joints. Clin Orthop 2000;379 Suppl:S17-S30.

36. Lieberman JR, Ghivizzani SC, Evans SC. Gene transfer approaches to the healing of bone and cartilage. Mol Ther 2002;6:141-147.

37. Viggeswarapu M Boden SD, Liu Y, et al. Adenoviral delivery of LIM mineralization protein-1 induces new-bone formation in vitro and in vivo. J Bone Joint Surg Am 2001;83:364-376.

38. Lieberman JR, Daluiski A, Einhorn TA. Current Concepts Review. The role of growth factors in the repair of bone: biology and clinical applications. J Bone Joint Surg Am 2002;84:1032-1044.

39. Linkart TA, Mohan S, Baylink DJ. Growth factors for bone growth and repair; IGF, TGF beta, and BMP. Bone 1996;19 (1 Suppl):1S-12S.

40. Vortkamp A, Pathi S, Peretti GM, et al. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 1998;71:65-76.

41. Le AX, Miclau T, Hu D, et al. Molecular aspects of healing in stabilized and non- stabilized fractures. J Orthop Res 2001;19:78-84.

42. Ferguson CM, Alpern E, Miclau T, et al. Does adult fracture repair recapitulate embryonic skeletal formation? Mech Dev 1999;87:57-66.

43. Thompson Z, Miclau T, Hu D, et al. A model for intramembranous ossification during fracture healing. J Orthop Res 2002;20:1091-1098.

44. Asano N, Yamakazi T, Seto M, et al. The expression and prognostic significance of bone morphogenetic protein-2 in patients with malignant fibrous histiocytoma. J Bone Joint Surg Br 2004;86: 607-612.

156

45. Guo W, Gorlick R, Ladanyi M, et al. Expression of bone morphogenetic proteins and receptors in sarcomas. Clin Orthop 1999;365: 175-183.

46. Robey PG, Young MF, Flanders KC, et al. Osteoblasts synthesize and respond to transforming growth factor-type beta (TGF-beta) in vitro. J Cell Biol 1987;105:457-463. 47. Joyce ME, Jingushi S, Bolander ME. Transforming growth factor-beta in the regulation of fracture repair. Orthop Clin North Am 1990;21:199-209.

48. Lind M, Schumaker B, Soballe K, et al. Transforming growth factor beta enhances fracture healing in rabbit tibiae. Acta Orthop Scand 1993;64:553-556.

49. Nielsen HM, Andreassen TT, Ledet T, et al. Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop Scand 1994;65:37-41.

50. Critchlow MA, Bland TS, Ashurst DE. The effect of exogenous transforming growth factor-beta 2 on healing fractures in the rabbit. Bone 1995;16:521-527.

51. Ozkaynak E, Rueger DC, Drier EA, et al. OP-1 cDNA encodes an osteogenic protein in the TGF-beta family. EMBO J 1990;2085-2093.

52. Wang EA, Rosen V, D’Alessandro JS, et al. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci U S A 1990;87:2220- 2224.

53. Reddi AH. Bone morphogenetic proteins: from basic science to clinical applications. J Bone Joint Surg Am 2001;83 Suppl 1, Part 1 S1:1-6.

54. Urist MR. Bone: Formation by autoinduction. Science 1965;150:893-899.

55. Wang EA, Rosen V, D’Alessandro JS, et al. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci U S A 1990;87:2220- 2224.

56. Wozney JM, Rosen V, Celeste AJ, et al. Novel regualators of bone formation: molecular clones and activities. Science 1988;242:1528-1534.

57. Celeste AJ, Iannazzi JA, Taylor RC, et al. Identification of transforming growth factor β family members present in bone-inductive protein purified from bovine bone. Proc Natl Acad Sci U S A 1990;87:9843-9847.

58. Heckman JD, Boyan BD, Aufdemorte TB, et al. The use of bone morphogenetic protein in the treatment of nonunion in a canine model. J Bone Joint Surg Am 1991;73:750-764.

157

59. Sciadini MF, Dawson JM, Johnson KD. Evaluation of bovine-derived bone protein with a natural coral carrier as a bone-graft substitute in a canine segmental defect model. J Orthop Res 1997;15:844-857.

60. Sciadini MF, Johnson KD. Evaluation of recombinant human bone morphogenetic protein-2 as a bone graft substitute in a canine segmental defect model. J Orthop Res 2000;18:289-302.

61. Pluhar GE, Manley PA, Heiner JP, et al. The effect of recombinant human bone morphogenetic protein-2 on femoral reconstruction with an intercalary allograft in a dog model. J Orthop Res 2001;19:308-317.

62. Zabka AG, Pluhar GE, Edwards RB III, et al. Histomorphometric description of allograft bone remodeling and union in a canine segmental femoral defect model: a comparison of rhBMP-2, cancellous bone graft, and absorbable collagen sponge. J Orthop Res 2001;19:318-327.

63. Muschler GF, Hyodo A, Manning T, et al. Evaluation of human bone morphogenetic protein 2 in a canine spinal fusion model. Clin Orthop 1994;308:229-240.

64. Sandhu HS, Kanim LEA, Kabo JM, et al. Evaluation of rhBMP-2 with an OPLA carrier in a canine posterolateral (transverse process) spinal fusion model. Spine 1995;20:2669-2682.

65. David SM, Gruber HE, Meyer RA Jr, et al. Lumbar spinal fusion using recombinant human bone morphogenetic protein in the canine. A comparison of three dosages and two carriers. Spine 1999;24:1973-1979.

66. Tatakis DN, Koh A, Jin L, et al. Peri-implant bone regeneration using recombinant human bone morphogenetic protein-2 in a canine model: a dose-response study. J Periodont Res 2002;37:93-100.

67. Sigurdsson TJ, Fu E, Tatakis DN, et al. Bone morphogenetic protein-2 for bone regeneration and osseointegration. Clin Oral Impl Res 1997;8:367-374.

68. Rodeo SA, Suzuki K, Deng XH, et al. Use of recombinant human bone morphogenetic protein-2 to enhance tendon healing in a bone tunnel. Am J Sports Med 1999;27:476-488.

69. Cook SD, Baffes GC, Wolfe MW, et al. Recombinant human bone morphogenetic protein-7 induces healing in a canine long-bone segmental defect model. Clin Orthop 1994;301:302-312.

158

70. Cook SD, Barrack RL, Santman M, et al. Strut allograft healing to the femur with recombinant human osteogenic protein-1. Clin Orthop 2000;381:47-57.

71. Salkeld SL, Patron LP, Barrack RL, et al. The effect of osteogenic protein-1 on the healing of segmental bone defects treated with autograft or allograft bone. J Bone Joint Surg Am 2001;83:803-816.

72. Mont MA, Jones LC, Elias JJ, et al. Strut-autografting with and without osteogenic protein-1. J Bone Joint Surg Am 2001;83:1013-1022.

73. Cullinane DM, Lietman SA, Inoue N, et al. The effect of recombinant human osteogenic protein-1 (bone morphogenetic protein-7) impregnation on allografts in a canine intercalary bone defect. J Orthop Res 2002;20:1240-1245.

74. Trippel SB. Growth factors as therapeutic agents. Instr Course Lect 1997;46:473-476.

75. Andrew JG, Hoyland J, Freemont AJ, et al. Insulin-like growth factor gene expression in human fracture callus. Calcif Tissue Int 1993;53:97-102.

76. Canalis E, McCarthy TL, Centrella M. Effects of platelet-derived growth factor on bone formation in vitro. J Cell Physiol 1989;140:530-537.

77. Bak B, Jorgensen PH, Andreassen TT. Dose response of growth hormone on fracture healing in the rat. Acta Orthop Scand 1990;61:54-57.

78. Bhaumick B, Bala RM. Differential effects of insulin-like growth factors I and II on growth, differentiation and glucoregulation in differentiating chondrocyte cells in culture. Acta Endocrinol (Copenh) 1991;125:201-211.

79. Trippel S, Wroblewski J, Makower A, et al. Regulation of growth-plate chondrocytes by insulin-like growth factor I and basic fibroblast growth factor. J Bone Joint Surg Am 1993;75:177-189.

80. Frisbie DD, Nixon AJ. Insulin-like growth factor 1 and corticosteroid modulation of chondrocyte metabolic and mitogenic activities in interleukin 1-conditioned equine cartilage. Am J Vet Res 1997;58:524-530.

81. Frisbie DD, Sandler EA, Trotter EA, et al. Metabolic and mitogenic activities of insulin-like growth factor-1 in interleukin-1-conditioned equine cartilage. Am J Vet Res 2000;61:436-441.

159

82. Fortier LA, Lust G, Mohammed HO, et al. Coordinate upregulation of cartilage matrix synthesis in fibrin cultures supplemented with exogenous insulin-like growth factor-I. J Orthop Res 1999;17:467-474.

83. Worster AA, Brower-Toland BD, Fortier LA, et al. Chondrocytic differentiation of mesenchymal stem cells sequentially exposed to transforming growth factor-beta1 in monolayer and insulin-like growth factor-I in a three-dimensional matrix. J Orthop Res 2001;19:738-749.

84. Fortier LA, Nixon AJ, Lust G. Phenotypic expression of equine articular chondrocytes grown in three-dimensional cultures supplemented with supraphysiologic concentrations of insulin-like growth factor-I. Am J Vet Res 2002;63:301-305.

85. Nixon AJ, Fortier LA, Williams J, et al. Enhanced repair of extensive articular cartilage defects by insulin-like growth factor-I-laden fibrin composites. J Orthop Res 1999;17:475-487.

86. Fortier LA, Balkman CE, Sandell LJ, et al. Insulin-like growth factor-I gene expression patterns during spontaneous repair of acute articular cartilage injury. J Orthop Res 2001;19:720-728.

87. Fortier LA, Mohammed HO, Lust G, et al. Insulin-like growth factor-I enhances cell- based repair of articular cartilage. J Bone Joint Surg Br 2002;84:276-288.

88. Fernihough JK, Innes JF, Billingham MEJ, et al. Changes in local regulation of insulin-like growth factors I and II and insulin-like growth factor binding proteins in osteoarthritis of the canine stifle joint secondary to cruciate ligament rupture. Vet Surg 2003;32:313-323.

89. Martel-Pelletier J, DiBattista JA, Lajeunesse D, et al. IGF/IGFBP axis in cartilage and bone in osteoarthritis pathogenesis. Inflamm Res 1998;47: 90-100.

91. Millis DL, Wilkens BE, Daniel GB, et al. Radiographic, densitometric, and biomechanical effects of recombinant canine somatotropin in an unstable ostectomy gap model of bone healing in dogs. Vet Surg 1998;27:85-93.

92. Vortkamp A, Lee K, Lanske B, et al. Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 1996;273:613-622.

93. Cho T-J, Gerstenfeld LC, Einhorn TA. Differential temporal expression of members of the transforming growth factor-β superfamily during murine fracture healing. J Bone Miner Res 2002 17:513-520. 160

94. Uusitalo H, Rantakokko J, Ahonen M, et al. A metaphyseal defect model of the femur for studies of murine bone healing. Bone 2001;28:423-429.

95. Vortkamp A, Pathi S, Peretti GM, et al. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 1998;71:65-76.

96. Iwasaki M, Le AX, Helms JA. Expression of Indian hedgehog, bone morphogenetic protein 6, and gli during skeletal morphogenesis. Mech Dev 1997;69:197-202.

97. Ferguson CM, Miclau T, Hu D, et al. Common molecular pathways in skeletal morphogenesis and repair. Ann NY Acad Sci 1998;857:33-42.

98. Le AX, Miclau T, Hu D, et al. Molecular aspects of healing in stabilized and non- stabilized fractures. J Orthop Res 2001;19:78-84.

99. Aldinger G, Herr G, Kusswetter W, et al. Bone morphogenetic protein: A review. Int Orthop 1991;15:169-177.

100. Wozney JM. The bone morphogenetic protein family and osteogenesis. Mol Reprod Dev 1992;32:160-167.

101. Boden SD. Clinical application of BMPs. Commentary. J Bone Joint Surg Am 2001;83 Suppl 1, Part 2, S161.

102. Govender S, Csimma C, Genant HK, et al. BMP-2 Evaluation in Surgery for Tibial Trauma (BESTT) Study Group. J Bone Joint Surg Am 2002;84:2123-2134.

103. Maniscalco P, Gambera D, Bertone C, et al. Healing of fresh tibial fractures with OP-1. A preliminary report. Acta Biomed Ataneo Parmense 2002;73:27-33.

104. Friedlaender GE, Perry GR, Cole JD, et al. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Joint Surg Am 2001;83 Suppl 1, Part 2:S151-S158.

105. Boden SD, Kang J, Sandhu H, et al. Use of recombinant bone morphogenetic protein-2 to achieve posterolateral lumbar spine fusion in humans: a prospective, randomized clinical pilot trial. Spine 2002;27:2662-2673.

106. Burkus JK, Transfeldt EE, Kitchel SH, et al. Clinical and radiographic outcomes of anterior lumbar interbody fusion using recombinant human bone morphogenetic protein- 2. Spine 2002;27:2396-2408.

161

107. Burkus JK, Dorchak JD, Sanders DL. Radiographic assessment of interbody fusion using recombinant human bone morphogenetic protein type 2. Spine 2003;28:372-377. 108. Johnsson R, Stromqvist B, Aspenberg P. Randomized radiostereoisomeric study comparing osteogenic protein-1 (BMP-7) and autograft bone in human noninstrumented posterolateral lumbar fusion. Spine 2002;27:2654-2661.

109. Martin GJ Jr, Boden SD, Marone MA, et al. Posterolateral intertransverse arthrodesis with rhBMP-2 in a nonhuman primate: important lessons learned regarding dose, carrier, and safety. J Spinal Disord 1999;12:179-186.

110. Wozney JM. Cellular and Molecular Biology of Bone. Bone morphogenetic proteins and their gene expression. Noda M, ed. New York, Academic Press, 1993:131-167.

111. Winn SR, Uludag H, Hollinger JO. Carrier systems for bone morphogenetic proteins. Clin Orthop 1999;367S:S95-S106.

112. Li RH, Bouxsein CA, Blake CA, et al. rh BMP-2 injected in a calcium phosphate paste (alpha-BSM) accelerates healing in the rabbit ulnar osteotomy model. J Orthop Res 2003;21:997-1004.

113. Marx RE, Carlson ER, Eichstaedt RM, et al. Platelet-rich plasma: growth factor enhancement for bone grafts. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85:638-646.

114. Furst G, Gruber R, Tangl S, et al. Sinus grafting with autogenous platelet-rich plasma and bovine hydroxyapatite. A histomorphometric study in minipigs. Clin Oral Implants Res 2003;14:500-508.

115. Kim SG, Kim WK, Park JC, et al. A comparative study of osseointegration of Avana implants in a demineralized freeze-dried bone alone or with platelet-rich plasma. J Oral Maxillofac Surg 2002;60:1018-1025.

116. Kim SG, Chung CH, Kim YK, et al. Use of particulate dentin-plaster of Paris combination with/without platelet-rich plasma in the treatment of bone defects around implants. Int J Oral Maxillofac Implants 2002;17:86-94.

117. Carter CA, Jolly DG, Worden CE, et al. Platelet-rich plasma gel promotes differentiation and regeneration during equine wound healing. Exp Mol Pathol 2003;74:244-255.

118. Yoo JU, Barthel TS, Nishimura K, et al. The chondrogenic potential of bone marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am 1998;80:1745- 1757. 162

119. Finkemeier CG. Current Concepts Review: Bone-grafting and bone-graft substitutes. J Bone Joint Surg Am 2002;84:454-464.

120. Devine MJ, Mierisch CM, Jang E, et al. Transplanted bone marrow cells localize to fracture callus in a mouse model. J Orthop Res 2002;20:1232-1239.

121. Tiedeman JJ, Connolly JF, Strates BS, et al. Treatment of nonunion by percutaneous injection of bone marrow and demineralized bone matrix. An experimental study in dogs. Clin Orthop 1991;268:294-302.

122. Connolly JF. Injectable bone marrow preparations to stimulate osteogenic repair. Clin Orthop 1995;313:8-18.

123. Connolly JF. Clinical use of marrow osteoprogenitor cells to stimulate osteogenesis. Clin Orthop 1998;355 Suppl:S257-266.

124. Kadiyala S, Kraus KH, Attawia MA, et al. Rapid bone regeneration in femoral defects by an autologous osteoprogenitor cell concentrate using an intraoperative selective cell retention technique, in Trans ORS 2003: Paper No. 0317.

125. Brodke DS, Kapur TA, Pedrozo HA, et al. A bone graft prepared with selective cell retention technology heals canine segmental defects as effectively as autograft. Available at www.cellectiveretention.com/cellect_files/obo9-60-000_canine_whitepaper.pdf. DePuy AcroMed, Inc., 2003. Accessed Jan. 15, 2004.

126. Frisbie DD, Ghivizzani SC, Robbins PD, et al. Treatment of experimental equine osteoarthritis by in vivo delivery of equine interleukin-1 receptor antagonist gene. Gene Ther 2002;9:12-20.

127. Bendele AM, Chlipala ES, Scherrer J, et al. Combination benefit of treatment with the cytokine inhibitors interleukin-1 receptor antagonist and PEGylated soluble tumor necrosis factor receptor type I in animal models of rheumatoid arthritis. Arthritis Rheum 2000;43:2648-2659.

128. Gouze J-N, Gouze E, Palmer GD, et al. A comparative study of the inhibitory effects of interleukin-1 receptor antagonist following administration as a recombinant protein or by gene transfer. Arthritis Res Ther 2003;5:R301-R309.

163

CHAPTER 2: GENE-MEDIATED DIFFERENTIATION OF STEM CELLS BY BMP-2 AND BMP-6

 1. ABI PRISM 7000 Sequence Detection System User Guide. Applied Biosystems, Foster City, CA 2002.

2. Adams S et al. Stimulation of type-X collagen gene transcription by retinoids occurs in part through the BMP signaling pathway. J Bone Joint Surg Am 2003;85:30-33.

  3. Affymetrix GeneChip Operating Software User’s Guide Version 1.2. Affymetrix, Inc. Santa Clara, CA, 2004.

4. Afrakhte M, Moren A, Jossan S, et al. Induction of inhibitory Smad6 and Smad7 mRNA by TGF-β family members. Biochem Biophys Res Commun 1998;249:505-11.

5. Ajenjo N, Canon E, Sanchez-Perez I, et al. Subcellular localization determines the protective effects of activated ERK2 against distinct apoptogenic stimuli in myeloid leukemia cells. J Biol Chem 2004; 279: 32813-23.

6. Alden TD, Pittman DD, Beres EJ, et al. Percutaneous spinal fusion using bone morphogenetic protein 2 gene therapy. J Neurosurg 1999;90:109-14.

7. Alden TD, Pittman DD, Hankins GR, et al. In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum Gene Ther 1999;10:2245-53.

8. Aoki H, Fujii M, Imamura T, et al. Synergistic effects of different bone morphogenetic protein type I receptors on alkaline phosphatase induction. J Cell Sci 2001;114:1483-89.

9. Aspenstrom P. The Rho GTPases have multiple effects on the actin cytoskeleton. Exp Cell Res 1999;246: 20-25.

10. Bagrodia S, Derijard B, Davis R, Cerione R. Cdc42 and PAK-mediated signaling leads to Jun kinase and p38 mitogen-activated protein kinase activation. J Biol Chem 1995;270:27995-98.

11. Baltzer AW, Lattermann C, Whalen JD, et al. Genetic enhancement of fracture repair: healing of an experimental segmental defect by adenoviral transfer of the BMP-2 gene. Gene Ther 2000;7:734-39.

164

12. Bellows CG, Heersche JN, Aubin JE. Determination of the capacity for proliferation and differentiation of osteoprogenitor cells in the presence and absence of dexamethasone. Dev Biol 1990;140:132-38.

13. Beresford JN. Osteogenic stem cells and the stromal system of bone and marrow. Clin Orthop 1989;240: 270-80.

14. Bertone AL, Pittman DD, Bouxsein ML et al. Adenoviral-mediated transfer of human BMP-6 gene accelerates healing in a rabbit ulnar osteotomy model. J Orthop Res 2004;22:1261-70.

15. Blum J, Parrott M, Mikos A, Barry M. Early osteoblastic differentiation induced by dexamethasone enhances adenoviral gene delivery to marrow stromal cells. J Orthop Res 2004;22:411-16.

16. Boskey A, Paschalis E, Binderman I, Doty S. BMP-6 accelerates both chondrogenesis and mineral maturation in differentiating chick limb-bud mesenchymal cell cultures. J Cell Biochem 2002;84:509-19.

17. Bruder SP, Jaiswal N, Haynesworth SE. Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J Cell Biochem 1997;64:278-94.

18. Bruder SP, Kraus KH, Goldberg VM, Kadiyala S. The effect of implants loaded with autologous mesenchymal stem cells on the healing of canine segmental bone defects. J Bone Joint Surg Am 1998;80:985-96.

19. Caplan AI, Bruder SP. Mesenchymal stem cells: building blocks for molecular medicine in the 21st century. Trends Mol Med 2001;7:259-64.

20. Celeste AJ, Iannazzi JA, Taylor RC, et al. Identification of transforming growth factor beta family members present in bone-inductive protein purified from bovine bone. Proc Natl Acad Sci U S A 1990;87:9843-47.

21. Chang SC-N, Chuang HL, Chen YR, et al. Ex vivo gene therapy in autologous bone marrow stromal cells for tissue-engineered maxillofacial bone regeneration. Gene Therapy 2003;10:2013-19.

22. Chang SC-N, Chuang H, Chen Y-R, et al. Cranial repair using BMP-2 gene engineering bone marrow stromal cells. J Surg Res 2004;119:85-91.

165

23. Cheng SL, Lou J, Wright NM, et al. In vitro and in vivo induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene. Calcif Tissue Int 2001;68:87-94.

24. Connolly JF. Injectable bone marrow preparations to stimulate osteogenic repair. Clin Orthop 1995;313:8-18.

25. Connolly JF, Guse R, Tiedeman J, Dehne R. Autologous marrow injection as a substitute for operative grafting of tibial nonunions. Clin Orthop 1991;266:259-70.

26. Cotter MJ, Muruve DA. The induction of inflammation by adenovirus vectors used for gene therapy. Front Biosci 2005;10:1098-1105.

27. DeBeer P, Schoenmakers EF, Twal WO, et al. The fibulin-1 gene (FBLN1) is disrupted in a t(12;22) associated with a complex type of synpolydactyly. J Med Genet 2002;39:98-104.

28. Dennis JE, Merriam A, Awadallah A, et al. A quadripotential mesenchymal progenitor cell isolated from the marrow of an adult mouse. J Bone Miner Res 1999;14:700-09.

29. Derubeis AR, Cancedda R. Bone marrow stromal cells (BMSCs) in bone engineering: limitations and recent advances. Ann Biomed Eng 2004;32:160-65.

30. Derynck R, Zhang Y, Feng XH. Smads: transcriptional activators of TGF-β responses. Cell 1998;95: 737-40.

31. Diduch DR, Jordan LCM, Mierisch CM, Balian G. Marrow stromal cells in alginate for repair of osteochondral defects. Arthroscopy 2000;16:571-77.

32. Engstrand, T, Daluiski A, Bahamonde ME, et al. Transient production of bone morphogenetic protein 2 by allogeneic transplanted transduced cells induces bone formation. Hum Gene Ther 2000;11:205-11.

33. Etienne-Manneville S, Hall A. Rho GTPases in cell biology. Nature 2002;12:62935.

34. Exton JH. Small GTPases minireview series. J Biol Chem 1999;273:19923.

35. Finkemeier CG. Bone-grafting and bone-graft substitutes. J Bone Joint Surg Am 2002;84:454-64.

166

36. Fortier LA, Deak MM, Semevolos SA, Cerione RA. Insulin-like growth factor-I diminishes the activation status and expression of the small GTPase Cdc42 in articular chondrocytes. J Orthop Res 2004;22: 436-45.

37. Fujii M, Takeda K, Imamura T, et al. Roles of bone morphogenetic protein type I receptors and Smad proteins in osteoblast and chondroblast differentiation. Mol Biol Cell 1999;10:3801-13.

38. Fukuda J, Nasu K, Sun B, et al. Expression of growth-regulated oncogene beta in an endometrial epithelial cell line, HHUA, and cultured human endometrial cells. J Reprod Immunol 2003;59:61-70.

39. Fukui N, Zhu Y, Maloney WJ, et al. Stimulation of BMP-2 expression by pro- inflammatory cytokines IL-1 and TNF-α in normal and osteoarthritic chondrocytes. J Bone Joint Surg Am 2003;85 Suppl 3:S59-S66.

40. Gazit D, Turgeman G, Kelley P, et al. Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone: a novel cell-mediated gene therapy. J Gene Med 1999;1:121-33.

41. Gelse K, von der Mark K, Aigner T, et al. Articular cartilage repair by gene therapy using growth factor-producing mesenchymal cells. Arthritis Rheum 2003;48:430-41.

 42. GeneChip Expression Analysis Technical Manual. Affymetrix, Inc. Santa Clara, CA, November 2004.

43. Grimsrud CD, Romano PR, D'Souza M, et al. BMP-6 is an autocrine stimulator of chondrocyte differentiation. J Bone Miner Res 1999;14:475-82.

44. Gu W, Bertone A. Generation and performance of an equine-specific large-scale gene expression microarray. Am J Vet Res 2004;65:1664-73.

45. Guo X, Rueger D, Higgins D. Osteogenic protein-1 and related bone morphogenetic proteins regulate dendritic growth and the expression of microtubule-associated protein-2 in rat sympathetic neurons. Neurosci Lett 1998;245: 131-34.

46. Hall A, Nobes CD. Rho GTPases: molecular switches that control the organization and dynamics of the actin cytoskeleton. Philos Trans R Soc Lond B 2000;355:965- 70.

47. Hanada K, Dennis JE, Caplan AI. Stimulatory effects of basic fibroblast growth factor and bone morphogenetic protein-2 on osteogenic differentiation of rat bone marrow-derived mesenchymal stem cells. J Bone Miner Res 1997;12:1606-14.

167

48. Hata A, Lagna G, Massagué J, Hemmati-Brivanlou A. Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev 1998;12:186-97.

49. Heckman JD, Ehler W, Brooks BP, et al. Bone morphogenetic protein but not transforming growth factor-beta enhances bone formation in canine diaphyseal nonunions with a biodegradable composite polymer. J Bone Joint Surg Am 1999;81:1717-29.

50. Horikoshi N, Cong J, Kley N, Shenk T. Isolation of differentially expressed cDNAs from p53-dependent apoptotic cells: activation of the human homologue of the Drosophila peroxidasin gene. Biochem Biophys Res Commun 1999;261:864-69.

51. Huang W, Rudkin GH, Carlsen B, et al. Overexpression of BMP-2 modulates morphology, growth, and gene expression in osteoblastic cells. Exp Cell Res 2002;274:226-34.

52. Imamura T, Takase M, Nishihara A, et al. Smad6 inhibits signaling by the TGF-β superfamily. Nature 1997; 389:622-26.

53. Izzo M, Pucci B, Tuan R, Hall D. Gene expression profiling following BMP-2 induction of mesenchymal chondrogenesis in vitro. Osteoarthritis Cartilage 2002;10:23-33.

54. Jane JA Jr, Dunford BA, Kron A, et al. Ectopic osteogenesis using adenoviral bone morphogenetic protein (BMP)-4 and BMP-6 gene transfer. Mol Ther 2002;6:464-70.

55. Johnstone B, Hering TM, Caplan AI, et al. In vitro chondrogenesis of bone marrow- derived mesenchymal progenitor cells. Exp Cell Res 1998;238:265-72.

56. Kanatani M, Sugimoto T, Kaji H, et al. Stimulatory effect of bone morphogenetic protein-2 on osteoclast-like cell formation and bone-resorbing activity. J Bone Miner Res 1995;10:1681-90.

57. Kavalkovich K, Boynton R, Murphy J, Barry F. Chondrogenic differentiation of human mesenchymal stem cells within an alginate layer culture system. In Vitro Cell Dev Biol Anim 2002;38:457-66.

58. Le AX, Miclau T, Hu D, Helms JA, et al. Molecular aspects of healing in stabilized and non-stabilized fractures. J Orthop Res 2001;19:78-84.

168

59. Lecanda F, Avioli LV, Cheng SL. Regulation of bone matrix protein expression and induction of differentiation of human osteoblasts and human bone marrow stromal cells by bone morphogenetic protein-2. J Cell Biochem 1997;67: 386-96.

60. Lieberman JR, Daluiski A, Stevenson S, et al. The effect of regional gene therapy with bone morphogenetic protein-2-producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 1999:81A:905-17.

61. Lieberman JR, Le LQ, Wu L, et al. Regional gene therapy with a BMP2-producing murine stromal cell line induces heterotopic and orthotopic bone formation in rodents. J Orthop Res 1998;16:330-39.

62. Lin X, Liang YY, Sun B, et al. Smad6 recruits transcription corepressor CtBP to repress bone morphogenetic protein-induced transcription. Mol Cell Biol 2003;23:9081-93.

63. Liu Q, Muruve DA. Molecular basis of the inflammatory response to adenovirus vectors. Gene Ther 2003;10:935-40.

64. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time -∆∆C quantitative PCR and the 2 T method. Methods 2001;25:402-08.

 65. Living Image Software Manual, Version 2.20, Xenogen Corporation, Alameda, CA, November 2002.

66. Lou J, Xu F, Merkel K, Manske P. Gene therapy: adenovirus-mediated human bone morphogenetic protein-2 gene transfer induces mesenchymal progenitor cell proliferation and differentiation in vitro and bone formation in vivo. J Orthop Res 1999;17:43-50.

67. Luria EA, Owen ME, Friedenstein AJ, et al. Bone formation in organ cultures of bone marrow. Cell Tissue Res 1987;248:449-54.

68. Ma H-L, Hung S-C, Lin S-Y, et al. Chondrogenesis of human mesenchymal stem cells encapsulated in alginate beads. J Biomed Mater Res 2002;64A:273-81.

69. Majumdar M, Wang E, Morris E. BMP-2 and BMP-9 promote chondrogenic differentiation of human multipotential mesenchymal cells and overcome the inhibitory effect of IL-1. J Cell Physiol 2001;189:275-84.

169

70. Majumdar M, Banks V, Peluso D, Morris E. Isolation, characterization, and chondrogenic potential of human bone marrow-derived multipotential stromal cells. J Cell Physiol 2000;185: 98-106.

71. Marcus J. The development and evolution of crossveins in insect wings. J Anat 2001;199:211-16.

72. Massagué J. How cells read TGF-beta signals. Nat Rev Mol Cell Biol 2000;1:169-78.

73. Mehta V, Kang Q, Luo J, et al. Characterization of adenovirus-mediated gene transfer in rabbit flexor tendons. J Hand Surg [Am] 2005;30:136-41.

74. Minina E, Wenzel HM, Kreschel C, et al. BMP and Ihh/PTHrP signaling interact to coordinate chondrocyte proliferation and differentiation. Development 2001;128:4523-24.

75. Mitchell M, Jerebtsova M, Batshaw ML, et al. Long-term gene transfer to mouse fetuses with recombinant adenovirus and adeno-associated virus (AAV) vectors. Gene Ther 2000;7:1986-92.

76. Moutsatsos IK, Turgeman G, Zhou S, et al. Exogenously regulated stem cell- mediated gene therapy for bone regeneration. Mol Ther 2001;3:449-61.

77. Mueller SM, Glowacki J. Age-related decline in the osteogenic potential of human bone marrow-derived mes9enchymal stem cells cultured in three-dimensional collagen sponges. J Cell Biochem 2001;82:583-90.

78. Musgrave DS, Bosch P, Ghivizzani S, et al. Adenovirus-mediated direct gene therapy with bone morphogenetic protein-2 produces bone. Bone 1999;24: 541-47.

79. Noël D, Gazit D, Bouquet C, et al. Short-term BMP-2 expression is sufficient for in vivo osteochondral differentiation of mesenchymal stem cells. Stem Cells 2004;22:74-85.

80. Oakes DA, Lieberman JR. Osteoinductive applications of regional gene therapy: ex vivo gene transfer. Clin Orthop 2000;379S:S101-S112.

81. Olmsted EA, Blum JS, Rill D, et al. Adenovirus-mediated BMP2 expression in human bone marrow stromal cells. J Cell Biochem 2001;82:11-21.

82. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143-47.

170

83. Porcellini A, Blasi AD. Viral infection protocols. Methods Mol Biol 2004;259:155- 66.

84. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997;276:71-74.

85. Reddi AH. Bone morphogenetic proteins, bone marrow stromal cells, and mesenchymal stem cells: Maureen Owen revisited. Clin Orthop 1995;313:115-19.

86. Sanceau J, Boyd D, Seiki M, Bauvois B. Interferons inhibit tumor necrosis factor- alpha-mediated matrix metalloproteinase-9 activation via interferon regulatory factor-1 binding competition with NF-kappa B. J Biol Chem 2002;277:35766-75.

87. Scaduto AA, Lieberman JR. Gene therapy for osteoinduction. Orthop Clin North Am 1999;30:625-33.

88. Sekiya I, Coulter D, Prockop D. BMP-6 enhances chondrogenesis in a subpopulation of human marrow stromal cells. Biochem Biophys Res Commun 2001;284:411-18.

89. Sekiya I, Larson BL, Vouristo JT, et al. Comparison of effect of BMP-2, -4, and -6 on in vitro cartilage formation of human adult stem cells from bone marrow stroma. Cell Tissue Res 2005 Mar 19; (Epub: http://www.springerlink.com/app/home/contribution.asp?wasp=770ed4e8d47a4f46a 071276c15508ab2&referrer=parent&backto=issue,12,19;journal,1,137;linkingpublic ationresults,1:100524,1)

90. Sekiya I, Vuoristo JT, Larson BL, et al. In vitro cartilage formation by human adult stem cells from bone marrow stroma defines the sequence of cellular and molecular events during chondrogenesis. Proc Natl Acad Sci USA 2002;99:4397-4402.

91. Simmons C, Alsberg E, Hsiong S, et al. Dual growth factor delivery and controlled scaffold degradation enhance in vivo bone formation by transplanted bone marrow stromal cells. Bone 2004;35:562-69.

92. Steinert A, Weber M, Dimmler A, et al. Chondrogenic differentiation of mesenchymal progenitor cells encapsulated in ultrahigh-viscosity alginate. J Orthop Res 2003;21:1090-97.

93. Takase M, Imamura T, Sampath TK, et al. Induction of Smad6 mRNA by bone morphogenetic proteins. Biochem Biophys Res Commun 1998;244:26-29.

94. Tavassoli M, Crosby WH. Transplantation of marrow to extramedullary sites. Science 1968;161:54-56. 171

95. Thies RS, Bauduy M, Ashton BA, et al. Recombinant human bone morphogenetic protein-2 induces osteoblastic differentiation in W-20-17 stromal cells. Endocrinology 1992;130:1318-24.

96. Wang J, Kanim LE, Yoo S, et al. Effect of regional gene therapy with bone morphogenetic protein-2-producing bone marrow cells on spinal fusion in rats. J Bone Joint Surg Am 2003;85:905-11.

97. Weber M, Steinert A, Jork A, et al. Formation of cartilage matrix proteins by BMP- transfected murine mesenchymal stem cells encapsulated in a novel class of alginates. Biomaterials 2002;23:2003-13.

98. Winter A, Breit S, Parsch D, et al. Cartilage-like gene expression in differentiated human stem cell spheroids. Arthritis Rheum 2003;48:418-29.

99. Wozney JM. The bone morphogenetic protein family and osteogenesis. Mol Reprod Dev 1992;32:160-67.

100. Yoo JU, Barthel TS, Nishimura K, et al. The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am 1998;80:1745-57.

101. Zakut R, Givol D. The tumor suppression function of p21 (Waf) is contained in its N-terminal half ('half-WAF'). Oncogene 1995;11:393-95.

102. Zhang LX, Mills KJ, Dawson MI, et al. Evidence for the involvement of retinoic acid receptor RAR alpha-dependent signaling pathway in the induction of tissue transglutaminase and apoptosis by retinoids. J Biol Chem 1995;270:6022-29.

CHAPTER 3: RODENT MODELS FOR THE STUDY OF ARTICULAR FRACTURE HEALING

1. Marsh JL, Buckwalter J, Gelberman R, et al. Articular fractures: does an anatomic reduction really change the result? J Bone Joint Surg Am. 2002;84:1259-1271.

2. Buckwalter JA, Mankin HJ. Instructional Course Lecture, the American Academy of Orthopaedic Surgeons. Articular cartilage. Part II: Degeneration and osteoarthrosis, repair, regeneration and transplantation. J Bone Joint Surg Am. 1997;79:612-632. 172

3. Buckwalter J, Rosenberg L, Coutts R, Hunziker E, Reddi A, Mow V: Articular cartilage: injury and repair. In: Injury and repair of the musculoskeletal soft tissues, pp 465-482. Ed by S Woo and J Buckwalter. Park Ridge, IL, American Academy of Orthopaedic Surgeons, 1988.

4. Buckwalter J, Rosenberg L, Hunziker E: Articular cartilage: composition, structure, response to injury, and methods of facilitating repair. In: Articular cartilage and knee joint function: basic science and arthroscopy, pp 19-56. Ed by J Ewing. New York, Raven, 1990.

5. Buckwalter J: Mechanical injuries of articular cartilage. In: Biology and biomechanics of the traumatized synovial joint: the knee as a model, pp 83-96. Ed by Finerman G and Noyes F. Rosemont, IL, American Academy of Orthopaedic Surgeons, 1992.

6. Bonnarens F, Einhorn T: Production of a standard closed fracture in laboratory animal bone. J Orthop Res 2:97-101, 1984.

7. Bourque W, Gross M, Hall B: A reproducible model for producing and quantifying the stages of fracture repair. Lab Anim Sci 42:369-374, 1992.

8. Hiltunen A, Vuorio E, Aro H: A standardized experimental fracture in the mouse tibia. J Orthop Res 11:305-312, 1993.

9. Nakase T, Nomura S, Yoshikawa H, et al.: Transient and localized expression of bone morphogenetic protein 4 messenger RNA during fracture healing. J Bone Miner Res 9:651-659, 1994.

10. Turgeman G, Pittman D, Muller R, et al.: Engineered human mesenchymal stem cells: a novel platform for skeletal cell mediated gene therapy. J Gene Med 3:240- 251, 2001.

11. Tomatsu T, Imai N, Takeuchi N, et al.: Experimentally produced fractures of articular cartilage and bone. J Bone Joint Surg Br 74:457-462, 1992.

12. Llinas A, McKellop H, Marshall G, et al.: Healing and remodeling of articular incongruities in a rabbit fracture model. J Bone Joint Surg Am 75:1508-1523, 1993.

13. Reuter J, Ovadia S, Howell P, Jaskwich D: Femoral fracture repair and postoperative management in New Zealand white rabbits. Contemp Top Lab Anim Sci 41:49-52, 2002.

14. Zachos TA, Bertone AL. Gene-mediated osteogenic differentiation of stem cells by bone morphogenetic proteins -2 or -6, J Orthop Res In Press, 2006. 173

15. Bertone AL, Pittman DD, Bouxsein ML et al. Adenoviral-mediated transfer of human BMP-6 gene accelerates healing in a rabbit ulnar osteotomy model. J Orthop Res 22:1261-1270, 2004.

16. Majumdar, MK, Banks V, Peluso DP, Morris EA. Isolation, characterization, and chondrogenic potential of human bone marrow-derived multipotent stromal cells. J Cell Physiol 185: 98-106, 2000.

17. Beamer W, Donahue L, Rosen C, Baylink D: Genetic variability in adult bone density among inbred strains of mice. Bone 18:397-403, 1996.

18. Judex S, Donahue L-R, Rubin C: Genetic predisposition to low bone mass is paralleled by an enhanced sensitivity to signals anabolic to the skeleton. FASEB J 16, 2002.

19. Boskey A, Wright T, van der Meulen MCH: Editorial: Guidelines for describing mouse skeletal phenotype. J Orthop Res 21:1-5, 2003.

20. deBoer J, van Blitterswijk C, Löwik C. Bioluminescent imaging: emerging technology for non-invasive imaging of bone tissue engineering. Biomaterials 27: 1851-1858, 2006. 21. Hollister SJ, Lin CY, Saito E, et al. Engineering craniofacial scaffolds. Orthod Craniofacial Res 8: 162-173, 2005.

22. Kohler T, Beyeler M, Webster D, et al. Compartmental bone morphometry in the mouse femur: reproducibility and resolution dependence of microtomographic measurements. Calcif Tissue Int 77: 281-290,2005.

23. Gabet Y, Müller R, Regev E et al. Osteogenic growth peptide modulates fracture callus structural and mechanical properties. Bone 35: 65-73, 2004.

24. Anastasiou, A and Hall LD. A novel RF coil configuration for in-vivo and ex-vivo imaging of arthritic rabbit knee joints. Magn Reson Imaging 21: 61-68, 2003.

174

CHAPTER 4: CHONDRO-OSSEOUS DIFFERENTIATION OF BONE MARROW- DERIVED MESENCHYMAL STEM CELLS IN ALGINATE CULTURES INDUCED BY BMP-2 AND BMP-6 GENE DELIVERY

1. Andrades JA, Santamaria JA, Nimni ME, Becerra J. Selection and amplification of a bone marrow cell population and its induction to the chondro-osteogenic lineage by rhOP-1: an in vitro and in vivo study. Int J Dev Biol 2001; 45: 689-693.

2. Sekiya I, Colter D, Prockop DJ. BMP-6 enhances chondrogenesis in a subpopulation of human marrow stromal cells. Biochem Biophys Res Commun 2001; 284: 411-418.

3. Sekiya I, Vuoristo J, Larson BL, Prockop DJ. In vitro cartilage formation by human adult stem cells from bone marrow stroma defines the sequence of cellular and molecular events during chondrogenesis. Proc Natl Acad Sci U S A 2002; 99: 4397- 4402.

4. Andrades JA, Han B, Nimni ME, Ertl DC, Simpkins RJ, Arrabal MP et al. A modified rhTGF-β1 and rhBMP-2 are effective in initiating a chondro-osseous differentiation pathway in bone marrow cells cultured in vitro. Connect Tissue Res 2003; 44: 188- 197.

5. Muraglia A, Corsi A, Riminucci M, Mastrogiacomo M, Cancedda R, Bianco P et al. Formation of a chondro-osseous rudiment in micromass cultures of human bone- marrow stromal cells. J Cell Sci 2003; 116: 2949-2955.

6. Noël D, Gazit D, Bouquet C, Apparailly F, Bony C, Plence P et al. Short-term BMP-2 expression is sufficient for in vivo osteochondral differentiation of mesenchymal stem cells. Stem Cells 2004; 22: 74-85.

7. Ichinose S, Tagami M, Muneta T, Sekiya I. Morphological examination during in vitro cartilage formation by human mesenchymal stem cells. Cell Tissue Res 2005; 322: 217-226.

8. Ichinose S, Yamagata K, Sekiya I, Muneta T, Takagami M. Detailed examination of cartilage formation and endochondral ossification using human mesenchymal stem cells. Clin Exp Pharmacol Physiol 2005; 32: 561-570.

9. Palmer GD, Steinert A, Pascher A, Gouze E, Gouze J-N, Betz O et al. Gene-induced chondrogenesis of primary mesenchymal stem cells in vitro. Mol Ther 2005; 12: 219- 228.

175

10. Sekiya I, Larson BL, Vuoristo JT, Reger RL, Prockop DJ. Comparison of effect of BMP-2, -4, and -6, on in vitro cartilage formation of human adult stem cells from bone marrow stroma. Cell Tissue Res 2005; 320: 269-276.

11. Majumdar MK, Banks V, Peluso DP, Morris EA. Isolation, characterization, and chondrogenic potential of human bone marrow-derived multipotent stromal cells. J Cell Physiol 2000; 185: 98-106.

12. Majumdar MK, Wang E, Morris EA. BMP-2 and BMP-9 promote chondrogenic differentiation of human multipotential mesenchymal cells and overcome the inhibitory effect of IL-1. J Cell Physiol 2001; 189: 275-284.

13. Kavalkovich KW, Boynton RE, Murphy JM, Barry F. Chondrogenic differentiation of human mesenchymal stem cells within an alginate layer culture system. In Vitro Cell Dev Biol Anim 2002; 38: 457-466.

14. Ma H-L, Hung S-C, Lin S-Y, Chen Y-L, Lo W-H. Chondrogenesis of human mesenchymal stem cells encapsulated in alginate beads. J Biomed Mater Res 2003; 64A: 273-281.

15. Weber M, Steinert A, Jork A, Dimmler A, Thurmer F, Schutze N, et al. Formation of cartilage matrix proteins by BMP-transfected murine mesenchymal stem cells encapsulated in a novel class of alginates. Biomaterials 2002; 23: 2003-2013.

16. Steinert A, Weber M, Dimmler A, Julius C, Schutze N, Noth U, et al. Chondrogenic differentiation of mesenchymal progenitor cells encapsulated in ultrahigh-viscosity alginate. J Orthop Res 2003; 21: 1090-1097.

17. Diduch DR, Jordan LCM, Mierisch CM, Balian G. Marrow stromal cells embedded in alginate for repair of osteochondral defects. Arthroscopy 2000; 16: 571-577.

18. Zilberman Y, Turgeman G, Pelled G, Nong X, Moutsatsos I, Hortelano G, et al. Polymer-encapsulated engineered adult mesenchymal stem cells secrete exogenously regulated rhBMP-2, and induce osteogenic and angiogenic tissue formation. Polym Adv Technol 2002; 13: 863-870.

19. Simmons CA, Alsberg E, Hsiong S, Kim WJ, Mooney DJ. Dual growth factor delivery and controlled scaffold degradation enhance in vivo bone formation by transplanted bone marrow stromal cells. Bone 2004; 35: 562-569.

20. Chang SC-N, Chuang H, Chen Y-R, Yang L-C, Chen JK, Mardini S, et al. Cranial repair using BMP-2 gene engineered bone marrow stromal cells. J Surg Res 2004 119: 85-91.

176

21. Edwards PC, Ruggiero S, Fantasia J, Burakoff R, Moorji SM, Paric E et al. Sonic hedgehog gene-enhanced tissue engineering for bone regeneration. Gene Ther 2005; 12: 75-86.

22. Mason JM, Grande DA, Barda M, Grant R, Pergolizzi RG, Breitbart AS. Expression of human bone morphogenetic protein 7 in primary rabbit periosteal cells: potential utility in gene therapy for osteochondral repair. Gene Ther 1998; 5: 1098-1104.

23. Alden TD, Pittman DD, Hankins GR, Beres EJ, Engh JA, Das S et al. In vivo endochondral bone formation using a bone morphogenetic protein 2 adenoviral vector. Hum Gene Ther 1999; 10: 2245-2253.

24. Cheng S-L, Lou J, Wright NM, Lai C-F, Avioli LV, Riew KD. In vitro and in vivo induction of bone formation using a recombinant adenoviral vector carrying the human BMP-2 gene. Calcif Tissue Int 2001; 68: 87-94.

25. Abe N, Lee Y-P, Sato M, Zhang X, Wu J, Mitani K et al. Enhancement of bone repair with a helper-dependent transfer of bone morphogenetic protein-2. Biochem Biophys Res Commun 2002; 297: 523-527.

26. Howard D, Partridge K, Yang X, Clarke NMP, Okubo Y, Bessho K et al. Immunoselection and adenoviral genetic modulation of human osteoprogenitors: in vivo bone formation on PLA scaffold. Biochem Biophys Res Commun 2002; 299: 208-215.

27. Olmsted-Davis EA, Gugala Z, Gannon FH, Yotnda P, McAlhany RE, Lindsey RW et al. Use of a chimeric adenovirus vector enhances BMP2 production and bone formation. Hum Gene Ther 2002; 13: 1337-1347.

28. Chang S-N, Chuang HL, Chen YR, Chen JK, Chung H-Y, Lu Y-L et al. Ex vivo gene therapy in autologous bone marrow stromal cells for tissue-engineered maxillofacial bone regeneration. Gene Ther 2003; 10: 2013-2019.

29. Gugala Z, Olmsted-Davis EA, Gannon FH, Lindsey RW, Davis AR. Osteoinduction by ex vivo adenovirus-mediated BMP2 delivery is independent of cell type. Gene Ther 2003; 10: 1289-1296.

30. Sugiyama O, Orimo H, Suzuki S, Yamashita K, Ito H, Shimada T. Bone formation following transplantation of genetically modified primary bone marrow stromal cells. J Orthop Res 2003; 21: 630-637.

31. Riew KD, Lou J, Wright NM, Cheng S-L, Bae KT, Avioli L. Thoracoscopic intradiscal spine fusion using a minimally invasive gene-therapy technique. J Bone Joint Surg Am 2003; 85: 866-871. 177

32. Wang JC, Kanim LEA, Yoo S, Campbell PA, Berk AJ, Lieberman JR. Effect of regional gene therapy with bone morphogenetic protein-2-producing bone marrow cells on spinal fusion in rats. J Bone Joint Surg Am 2003; 85: 905-911.

33. Schreiber RE, Blease K, Ambrosio A, Amburn E, Sosnowski B, Sampath TK. Bone induction by AdBMP-2/collagen implants. J Bone Joint Surg Am 2005; 87: 1059- 1068.

34. Bertone AL, Pittman DD, Bouxsein ML, Li J, Seeherman HJ. Adenoviral-mediated transfer of human BMP-6 gene accelerates healing in a rabbit ulnar osteotomy model. J Orthop Res 2004; 22: 1261-1270.

35. Jane JA Jr, Dunford BA, Kron A, Pittman DD, Sasaki T, Li JZ et al. Ectopic osteogenesis using adenoviral bone morphogenetic protein (BMP)-4 and BMP-6 gene transfer. Mol Ther 2002; 6: 464-470.

36. Yoo JU, Barthel TS, Nishimura K, Solchaga L, Caplan AI, Goldberg VM et al. The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am 1998; 80: 1745-1757.

37. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999; 284: 143-147.

38. Lieberman JR, Daluiski A, Stevenson S, Wu L, McAllister P, Lee YP et al. The effect of regional gene therapy with bone morphogenetic protein-2-producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 1999; 81: 905-917.

40. Worster, AA, Brower-Toland BD, Fortier LA, Bent SJ, Williams J, Nixon AJ. Chondrocytic differentiation of mesenchymal stem cells sequentially exposed to transforming growth factor-beta 1 in monolayer and insulin-like growth factor-I in a three-dimensional matrix. J Orthop Res 2001; 19: 738-749.

41. Huang CY, Reuben PM, D’Ipplolito G, Schiller PC, Cheung HS. Chondrogenesis of human bone marrow-derived mesenchymal stem cells in agarose culture. Anat Rec A Discov Mol Cell Evol Biol 2004; 378: 428-436.

42. Mauck RL, Yuan X, Tuan RS. Chondrogenic differentiation and functional maturation of bovine mesenchymal stem cells in long-term agarose culture. Osteoarthritis Cartilage 2006; 14: 179-189.

178

43. Celeste AJ, Iannazzi JA, Taylor RC, Hewick RM, Rosen V, Wang EA, et al. Identification of transforming growth factor beta family members present in bone- inductive protein purified from bovine bone. Proc Natl Acad Sci U S A 1990; 87: 9843-9847.

44. Zachos TA, Bertone AL. Gene-mediated osteogenic differentiation of stem cells by bone morphogenetic proteins-2 or -6. J Orthop Res 2006; In press.

45. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time -∆∆C quantitative PCR and the 2 T method. Methods 2001; 25: 402-408.

46. Park Y, Sugimoto M, Watrin A, Chiquet M, Hunziker EB. BMP-2 induces the expression of chondrocyte-specific genes in three-dimensional alginate hydrogel. Osteoarthritis Cartilage 2005; 13: 527-536.

CHAPTER 5: MESENCHYMAL STEM CELL-MEDIATED GENE DELIVERY OF BMP-2 IN AN ARTICULAR FRACTURE MODEL

1. Murray, M.M., Zurakowski, D., Vrahas, M.S. (2004). The death of articular chondrocytes after intra-articular fracture in humans. J Trauma 56: 128-131.

2. Marsh, J.L., et al. (2002). Articular fractures: does an anatomic reduction really change the result? J Bone Joint Surg Am. 84: 1259-1271.

3. Buckwalter, J.A. and Mankin, H.J. (1997) Instructional Course Lecture, the American Academy of Orthopaedic Surgeons. Articular cartilage. Part II: Degeneration and osteoarthrosis, repair, regeneration and transplantation. J Bone Joint Surg Am. 79: 612- 632.

4. Buckwalter, J.A., Rosenberg, L., Coutts, R., Hunziker, E, Reddi, A., and Mow, V. (1988). Articular cartilage: injury and repair. In: Injury and repair of the musculoskeletal soft tissues (S. Woo and J. Buckwalter, Eds.), pp. 465-482. American Academy of Orthopaedic Surgeons, Park Ridge, IL.

5. Buckwalter J.A., Rosenberg L., and Hunziker E. (1990). Articular cartilage: composition, structure, response to injury, and methods of facilitating repair. In Articular cartilage and knee joint function: basic science and arthroscopy. (J Ewing, Ed.) pp 19-56. Raven New York. 179

6. Buckwalter, J.A. (1992). Mechanical injuries of articular cartilage. In Biology and biomechanics of the traumatized synovial joint: the knee as a model (G. Finerman and F. Noyes, Eds.), pp. 83-96. American Academy of Orthopaedic Surgeons, Rosemont, IL.

7. Bonnarens, F. and Einhorn, T.A. (1984). Production of a standard closed fracture in laboratory animal bone. J Orthop Res 2: 97-101.

8. Bourque, W.T, Gross, M., and Hall, B.K. (1992). A reproducible model for producing and quantifying the stages of fracture repair. Lab Anim Sci 42: 369-374.

9. Hiltunen, A., Vuorio, E., and Aro, H.T. (1993). A standardized experimental fracture in the mouse tibia. J Orthop Res 11: 305-312.

10. Nakase, T., et al. (1994). Transient and localized expression of bone morphogenetic protein 4 messenger RNA during fracture healing. J Bone Miner Res 9: 651-659.

11. Turgeman, G., et al. (2001). Engineered human mesenchymal stem cells: a novel platform for skeletal cell mediated gene therapy. J Gene Med 3: 240-251.

12. Tomatsu, T., Imai, N., Takeuchi, N., Takahashi, K., and Kimura, N. (1992). Experimentally produced fractures of articular cartilage and bone. The effects of shear forces on the pig knee. J Bone Joint Surg Br 74: 457-462.

13. Llinas, A., McKellop, H.A., Marshall, G.J., Sharpe, F., Kirchen, M., and Sarmiento, A. (1993). Healing and remodeling of articular incongruities in a rabbit fracture model. J Bone Joint Surg Am 75:1508-1523.

14. Reuter, J.D., Ovadia, S., Howell, P., and Jaskwich, D.H. (2002). Femoral fracture repair and postoperative management in New Zealand white rabbits. Contemp Top Lab Anim Sci 41: 49-52.

15. Sekiya, I., Vuoristo, J., Larson, B.L., and Prockop, D.J. (2002). In vitro cartilage formation by human adult stem cells from bone marrow stroma defines the sequence of cellular and molecular events during chondrogenesis. Proc Natl Acad Sci U S A 99: 4397-4402.

16. Yoo, J.U., Barthel, T.S., Nishimura, K., Solchaga, L., Caplan, A.I., Goldberg, V.M., et al. (1998). The chondrogenic potential of human bone-marrow-derived mesenchymal progenitor cells. J Bone Joint Surg Am 80: 1745-1757.

17. Pittenger, M.F., Mackay, A.M., Beck, S.C., Jaiswal, R.K., Douglas, R., Mosca, J.D., et al. (1999). Multilineage potential of adult human mesenchymal stem cells. Science 284: 143-147. 180

18. Lieberman, J.R., Daluiski, A., Stevenson, S., Wu, L., McAllister, P., Lee, Y.P., et al. (1999). The effect of regional gene therapy with bone morphogenetic protein-2- producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 81: 905-917.

19. Andrades, J.A., Santamaria, J.A., Nimni, M.E., and Becerra, J. (2001). Selection and amplification of a bone marrow cell population and its induction to the chondro- osteogenic lineage by rhOP-1: an in vitro and in vivo study. Int J Dev Biol 45: 689- 693.

20. Andrades, J.A., Han, B., Nimni, M.E., Ertl, D.C., Simpkins, R.J., Arrabal, M.P., et al. (2003). A modified rhTGF- 1 and rhBMP-2 are effective in initiating a chondro- osseous differentiation pathway in bone marrow cells cultured in vitro. Connect Tissue Res 44: 188-197.

21. Muraglia, A., Corsi, A., Riminucci, M., Mastrogiacomo, M., Cancedda, R., Bianco, P., et al. (2003). Formation of a chondro-osseous rudiment in micromass cultures of human bone-marrow stromal cells. J Cell Sci 116: 2949-2955.

22. Noël, D., et al. (2004). Short-term BMP-2 expression is sufficient for in vivo osteochondral differentiation of mesenchymal stem cells. Stem Cells 22: 74-85.

23. Gazit, D., et al. (1999). Engineered pluripotent mesenchymal cells integrate and differentiate in regenerating bone: a novel cell-mediated gene therapy. J Gene Med 1:121-133.

24. Chang, S.C., et al. (2002). Ex vivo gene therapy in autologous bone marrow stromal cells for tissue-engineered maxillofacial bone regeneration. Gene Therapy 10: 2013- 2019.

25. Moutsatsos, I.K., et al. (2001). Exogenously regulated stem cell-mediated gene therapy for bone regeneration. Mol Ther 3:449-461.

26. Lou, J., Xu, F., Merkel, K., and Manske, P. (1999). Gene therapy: adenovirus- mediated human bone morphogenetic protein-2 gene transfer induces mesenchymal progenitor cell proliferation and differentiation in vitro and bone formation in vivo. J Orthop Res 17: 43-50.

27. Zachos, T.A. and Bertone, A.L. (2006) Gene-mediated osteogenic differentiation of stem cells with bone morphogenetic proteins-2 or -6. J Orthop Res, In Press.

28. Gnecchi, M., et al. (2006). Evidence supporting paracrine hypothesis for Akt- modified mesenchymal stem cell-mediated cardiac protection and functional improvement. FASEB J 20: 661-669. 181

29. Makkar, R.R., et al. (2005). Intramyocardial injection of allogenic bone marrow- derived mesenchymal stem cells without immunosuppression preserves cardiac function in a porcine model of myocardial infarction. J Cardiovasc Pharmacol Ther 10: 225-233.

30. Wang, Y., Haider, H.K., Ahmad, N., Xu, M., Ge, R., and Ashraf, M. (2006). Combining pharmacological mobilization with intramyocardial delivery of bone marrow cells over-expressing VEGF is more effective for cardiac repair. J Mol Cell Cardiol Apr 4 (E-pub ahead of print).

31. Lim, S.Y., et al. (2006). The effects of mesenchymal stem cells transduced with Akt in a porcine myocardial infarction model. Cardiovasc Res Mar 6 (Epub ahead of print).

32. Kim, D., et al. (2006). Retrovirus-mediated gene transfer of RANK-Fc prevents bone loss in ovariectomized mice. Stem Cells Mar 23 (Epub ahead of print).

33. Hauger, O., et al. (2006). MR evaluation of the glomerular homing of magnetically labeled mesenchymal stem cells in a rat model of nephropathy. Radiology 238: 200- 210.

34. Juncosa-Melvin, N., et al. (2006). The effect of autologous mesenchymal stem cells on the biomechanics and histology of gel-collagen sponge constructs used for rabbit patellar tendon repair. Tissue Eng 12: 369-379.

35. Hoffmann, A., et al. (2006). Neotendon formation induced by manipulation of the Smad8 signaling pathway in mesenchymal stem cells. J Clin Invest 116: 940-952.

36. Devine, M.J., Mierisch, C.M., Jang, E., Anderson, P.C., Balian, G.. (2002). Transplanted bone marrow cells localize to fracture callus in a mouse model. J Orthop Res 20: 1232-1239.

37. Majumdar, M.K., Banks, V., Peluso, D.P., and Morris, E.A. (2000). Isolation, characterization, and chondrogenic potential of human bone marrow-derived multipotent stromal cells. J Cell Physiol 185: 98-106.

38. Majumdar M.K., Wang, E., and Morris, E.A. (2001). BMP-2 and BMP-9 promote chondrogenic differentiation of human multipotential mesenchymal cells and overcome the inhibitory effect of IL-1. J Cell Physiol 189: 275-284. 39. Kavalkovich, K.W., Boynton, R.E., Murphy, J.M., and Barry, F. (2002). Chondrogenic differentiation of human mesenchymal stem cells within an alginate layer culture system. In Vitro Cell Dev Biol Anim 38: 457-466.

182

40. Ma, H.-L., Hung, S.-C., Lin, S.-Y., Chen, Y.-L., and Lo, W.-H. (2003). Chondrogenesis of human mesenchymal stem cells encapsulated in alginate beads. J Biomed Mater Res 64A: 273-281.

41. Weber, M., Steinert, A., Jork, A., Dimmler, A., Thurmer, F., Schutze, N., et al. (2002). Formation of cartilage matrix proteins by BMP-transfected murine mesenchymal stem cells encapsulated in a novel class of alginates. Biomaterials 23: 2003-2013.

42. Steinert, A., Weber, M., Dimmler, A., Julius, C., Schutze, N., Noth, U., et al. (2003). Chondrogenic differentiation of mesenchymal progenitor cells encapsulated in ultrahigh-viscosity alginate. J Orthop Res 21: 1090-1097.

43. Diduch, D.R., Jordan, L.C.M., Mierisch, C.M., and Balian, G. (2000). Marrow stromal cells embedded in alginate for repair of osteochondral defects. Arthroscopy 16: 571-577. 44. Zilberman, Y., Turgeman, G., Pelled, G., Nong, X., Moutsatsos, I., Hortelano, G. et al. (2002). Polymer-encapsulated engineered adult mesenchymal stem cells secrete exogenously regulated rhBMP-2, and induce osteogenic and angiogenic tissue formation. Polym Adv Technol 13: 863-870.

45. Simmons, C.A., Alsberg, E., Hsiong, S., Kim, W.J., and Mooney, D.J. (2004). Dual growth factor delivery and controlled scaffold degradation enhance in vivo bone formation by transplanted bone marrow stromal cells. Bone 35: 562-569.

46. Chang, S.C.-N., Chuang, H., Chen, Y.-R., Yang, L.-C., Chen, J.K., Mardini, S., et al. (2004). Cranial repair using BMP-2 gene engineered bone marrow stromal cells. J Surg Res 119: 85-91.

47. Lieberman, J.R., et al. (1998). Regional gene therapy with a BMP-2-producing murine stromal cell line induces heterotopic and orthotopic bone formation in rodents. J Orthop Res 16: 330-339.

48. Jane, J.A. Jr., et al. (2002). Ectopic osteogenesis using adenoviral bone morphogenetic protein (BMP)-4 and BMP-6 gene transfer. Mol Ther 6: 464-470.

49. Bertone, A.L., Pittman, D.D., Bouxsein, M.L., Li, J., Clancy, B., and Seeherman, H.J. (2004). Adenoviral-mediated transfer of human BMP-6 gene accelerates healing in a rabbit ulnar osteotomy model. J Orthop Res 22: 1261-1270.

50. Baltzer, A.W., et al. (2000). Genetic enhancement of fracture repair: healing of an experimental segmental defect by adenoviral transfer of the BMP-2 gene. Gene Ther 7: 734-739.

183

51. Betz, O.B., et al. (2006). Direct percutaneous gene delivery to enhance healing of segmental bone defects. J Bone Joint Surg Am 88: 355-365.

52. Egermann, M., et al. (2006). Direct adenoviral transfer of bone morphogenetic protein-2 cDNA enhances fracture healing in osteoporotic sheep. Hum Gene Ther Apr 5 (Epub ahead of print).

53. Southwood, L.L., Frisbie, D.D., Kawcak, C.E. Ghivizzani, S.C., Evans, C.H., and McIlwraith, C.W. (2004). Evaluation of Ad-BMP-2 for enhancing fracture healing in an infected defect fracture rabbit model. J Orthop Res 22: 66-72.

54. Sampath, T.K. and Reddi, A.H. (1981). Dissociative extraction and reconstruction of extracellular matrix components involved in local bone differentiation. Proc Natl Acad Sci U S A 78: 7599-7603.

55. Yamamoto, M., Tabata, Y., and Ikada, Y. (1998). Ectopic bone formation induced by biodegradable hydrogels incorporating bone morphogenetic protein. J Biomater Sci Polymer Ed 9: 439-458.

56. Takita, H., et al. (2004). Carrier dependent cell differentiation of bone morphogenetic protein-2 induced osteogenesis and chondrogenesis during the early implantation stage in rats. J Biomed Mater Res 71A: 181-189.

57. Bar, I., et al. (2003). Molecular imaging of the skeleton: quantitative real-time bioluminescence monitoring gene expression in bone repair and development. J Bone Miner Res 18: 570-578. 58. Gafni, Y., et al. (2004). Gene therapy platform for bone regeneration using an exogenously regulated, AAV-2-based gene expression system. Mol Ther 9: 587-595.

59. Celeste, A.J., et al. (1990). Identification of transforming growth factor beta family members present in bone-inductive protein purified from bovine bone. Proc Natl Acad Sci U S A 87: 9843-9847.

60. Erben, R. (1997). Embedding of bone samples in methylmethacrylate: an improved method suitable for bone histomorphometry, histochemistry, and immunohistochemistry. J Histochem Cytochem 45: 307-313.

61. Rohde, C., Anderson, D.E., Bertone, A.L., Weisbrode, S.E. (2000). Effects of phenylbutazone on bone activity and formation in horses. Am J Vet Res 61: 537-543.

62. Hupel, T.M., Weinberg, J.A., Aksenov, S.A., Schemitsch, E.H. (2001). Effect of unreamed, limited reamed, and standard reamed intramedullary nailing on cortical bone porosity and new bone formation. J Orthop Trauma 15: 18-27.

184